Search tips
Search criteria 


Logo of bumcprocBaylor University Medical Center ProceedingsAbout the JournalBaylor Health Care SystemSubmit a Manuscript
Proc (Bayl Univ Med Cent). 2017 April; 30(2): 147–150.
PMCID: PMC5349808

Pre and post hoc analysis of electronic health record implementation on emergency department metrics


Longitudinal time-based emergency department (ED) performance measures were quantified 12 months before and 12 months after (March 2012–February 2014) implementation of a Meditech 6.0® electronic health record (EHR) at a single urban academic ED. Data assessed were length of stay from door to door, door to admission, door to bed, bed to provider, provider to disposition, and disposition to admission, as well as number of patients leaving against medical advice and number of patients leaving without being seen. Analysis of variance was used to compare levels before and after EHR implementation for each variable, with adjustments made for the number of admissions, transfers, and month. No difference was seen in monthly volume, admissions, or transfers. Implementation of an EHR resulted in a sustained increase in ED time metrics for mean length of stay and times from door to door, door to admission, door to bed, and provider to disposition. Decreased ED time metrics were seen in bed-to-provider and disposition-to-admit times. The number of patients who left against medical advice increased after implementation, but the number of patients who left without being seen was not significantly different. Thus, EHR implementation was associated with an increase in time with most performance metrics. Although general times trended back to near preimplementation baselines, most ED time metrics remained elevated beyond the study length of 12 months. Understanding the impact of EHR system implementation on the overall performance of an ED can help departments prepare for potential adverse effects of such systems on overall efficiency.

An increasing number of hospitals and health care centers are adopting electronic health record (EHR) systems with the goal of improving health care quality while potentially decreasing costs (1). However, there are concerns regarding how efficiency and physician productivity are affected secondary to EHR implementation. Unlike ambulatory and inpatient settings, where patient volume can be adjusted to help with this transition, EDs are unable to alter volume and must maximize efficiency during this process. Currently, limited data exist in the ED literature showing the effects on productivity and the length of those effects, and research has shown that errors and unanticipated problems will arise from implementation (2). In addition, while computer physician order entry (CPOE) can provide many benefits when orders need to be placed and processed quickly (3), its implementation has unanticipated adverse effects, such as workflow issues, difficulties in the transition away from paper records, increased system demands, overdependence on technology, and loss of professional autonomy (4), which may affect overall department efficiency, resulting in ED crowding (57) and patient elopement (8). Several crowding measures have been endorsed by the National Quality Forum and Joint Commission, such as ED length of stay (LOS), waiting times, disposition to admission times, and rates of patients leaving without being seen (9, 10), and soon hospitals will report ED crowding measures to the Centers for Medicare and Medicaid Services to receive full Medicare payment (11). To date, little to no data exist showing results for an extended period of time surrounding the implementation of an EHR system. The objectives of this study were 1) to describe the effects of EHR implementation on various ED-specific metrics over the course of 12 months and 2) to compare those metrics to the 12 months prior to EHR implementation.


This retrospective analysis of ED metrics was performed at a single urban, university-affiliated, public, 25-bed ED in Tulsa, Oklahoma, with an emergency medicine residency program and an annual ED census of roughly 46,000. The ED was staffed with board-certified emergency physicians, emergency medicine residents, and other rotating residents from various services (i.e., internal medicine, family medicine, surgery, etc.).

Prior to implementation, the ED utilized paper documentation sheets, dictation, and written physician order entry via a clerical tech. Meditech 6.0® was implemented in the ED in a stepwise manner on March 1, 2013, with implementation completed on May 7, 2013. The first step involved implementation for registration, medical records review, laboratory, radiograph results, and electronic ED tracker board. Physician orders continued on paper until May 7, 2013, when CPOE was implemented. A simultaneous project was implemented on March 3, 2013, as a bedding initiative in an effort to reduce the door-to-admission time for newly admitted patients. Resident physicians utilized paper documentation with hand-off to supervising attending physicians. Board-certified emergency medicine attending physicians utilized dictation for formal, electronic patient encounter documentation.

ED metrics were analyzed during the timeframe of March 1, 2012, to February 28, 2014, with the break in pre- and postimplementation occurring on March 1, 2013. This information was compiled on a data information sheet for each 24-hour day. Individual data points were collected through standardized reporting from the ED operations committee. The ED metrics were collected for each day and grouped as dependent variables; they included LOS for admitted patients and nonadmitted patients, door-to-door time for discharged patients, door-to-admission time, door-to-bed time, bed-to-provider time, provider-to-disposition time, and disposition-to-admit time. In addition to the service time metrics above, the data for total ED visits, admission rates, and transfer rates were collected. Outcome measures for patient flow that may correlate with prolonged ED service times were also measured and included leaving against medical advice and leaving without being seen. Each of these grouped dependent variables was used in analysis of variance and compared to the ED metric timeframes before and after EHR implementation for each metric. The three covariates used to adjust for month-to-month variation and patient acuity differences within the groups were month, admission rate, and transfer rate.

Excel was used to calculate time metrics. Considering that the data set contained errors, patient data were only excluded if the data resulted in the integer of “0.” When standard errors and means were calculated with and without these data points, there was no significant difference. To determine standard error between the two groups, the dependent variable time stamps were determined for each patient encounter. These were averaged each day, then for the month, and then adjusted as above for each of the three covariates. The time metrics were grouped into pre- and postimplementation. Means and standard errors of the two major groups for each dependent variable were obtained for the 12 months before and the 12 months after EHR implementation. P values were obtained by comparing pre vs. post and adjusting for number of admits, transfers, and month. Any P value <0.05 was considered statistically significant.


A total of 100,198 ED visits were reviewed and included in this analysis, including 701,323 unique ED metric data points. Of these metric data points, 378,560, or 54%, were from after EHR implementation. During the course of the study, 13,174 patients were admitted, with the remaining number of patients either being discharged from the ED, transferred to another facility, leaving against medical advice, or leaving without being seen. The average monthly volume did not significantly change after implementation (P = 0.11) (Figure). Similarly, the monthly admission and transfer rates were also not significantly different from pre- to postimplementation, each with a P value of 0.06 (Table).

Monthly census means.
Metrics before and after implementation of an electronic health record in an emergency department

The mean LOS increased from 92.4 to 95.4 minutes (P = 0.01), with a change that persisted for more than 12 months after implementation. This trend persisted through many of the service intervals. The mean door-to-door time for total throughput time of ambulatory patients remained prolonged by increasing from 76.8 minutes to 81.6 minutes (P = 0.01). The mean door-to-bed time showed a statistically significant increase from 10.8 to 13.8 minutes. The mean provider-to-disposition time increased from 48.0 to 49.8 minutes (P = 0.01). Door-to-admission time was not significantly different between pre- and postimplementation (Table).

Even though many of the service intervals were prolonged, there were improvements after implementation for two separate timeframes. The first of these was for bed-to-provider time. This service interval was shortened by nearly 1.5 minutes, from 4.2 minutes to 3 minutes (P < 0.01). The second service interval that improved in average time was the disposition-to-admit, which improved from 85.8 to 78.6 minutes, a mean difference of 7.2 minutes. It should be noted, however, that the P value of 0.8 implied that this was not a statistically significant difference (Table).

The two clinical outcome measures of leaving without being seen and leaving against medical advice had surprisingly different significance between the two groups. The number leaving against medical advice nearly doubled from 6.3 (SE 0.85) patients per month to 11.6 (SE 1.17) per month (P < 0.01), yet the number leaving without being seen remained similar at 19.5 patients per month before implementation and 15.5 patients per month after implementation (P = 0.24) (Table).


Many facilities struggle to manage the same volume and acuity of patients in the same timely manner as they had prior to EHR implementation. This study has added an additional purview of similar results, with the addition of a longer data collection timeframe. Overall, patient visit metrics appeared to be mostly negatively impacted during the EHR implementation. LOS and door-to-door, door-to-bed, and provider-to-disposition times were all found to be longer after implementation, yet improvements in bed-to-provider and disposition-to-admit times after EHR implementation were surprising.

The first service time noted to have a trend toward improvement was disposition to admit. Soon after EHR implementation, this metric was noted to be a large component of the overall LOS. The ED had challenges with moving patients who have a disposition for admission to an inpatient hospital bed in a timely manner. A departmental goal was implemented in March 2013 to decrease disposition-to-admit times to a target of <45 minutes once disposition for admission was determined by the provider. This effort to improve performance likely resulted in a shortened disposition-to-admit time. Therefore, it remains unclear what effect EHR had in improving this metric.

The second improved timeframe was bed to provider, which decreased by nearly 1.5 minutes. It is likely, however, that this finding was a result of a change in procedure. Prior to EHR implementation, bed-to-provider service times were taken from providers' documentation of their start time on a paper documentation sheet in the room with the paper medical chart. However, once EHR was implemented, the bed-to-provider time was initiated when the provider signed up for the patient on the computer screen. Additionally, it has been observed that providers frequently initiate patient encounters prior to their registration in the computer. This could result in the provider completing a history, physical, and perhaps early electrocardiogram or I-Stat evaluation prior to initiating the mouse click in the computer, which registers the bed-to-provider time. These elements are likely the cause for the differences in service time and may have resulted in an artificially lowered time metric. The inconsistencies in documenting bed-to-provider time portend unreliable data analysis. More research is needed on this metric to make a more definitive conclusion regarding EHR's effect on it.

The negative impact from EHR implementation was seen in most of the metrics when comparing year-to-year data. With significant and trended increases in LOS, as well as door-to-door, door-to-admission, door-to-bed, and provider-to-disposition times, implementation of EHR had a primarily negative impact on the ED throughput metrics and service times over a 12-month period. Similarly to the study of Ward et al (5), ED physicians described themselves wading through patient encounters with cumbersome, disjointed movements. Once user and operations knowledge improved, this began to ease.

Many EHR implementations are all or nothing—i.e., they are all-encompassing and include medical records/chart reviewing, CPOE, documentation, and disposition paperwork (discharge instructions and prescriptions). At the study institution, a staged approach was employed. The EHR hospital system went live on March 1, 2013. The CPOE implementation was delayed for the ED until May 7, 2013. The hospital-wide CPOE went live on April 14, 2014. This was also a time when admission rates increased. Furthermore, the ED's documentation method had little to no change between paper documentation of a chart to dictation pre- and postimplementation. In many EHR implementations, a change in the documentation process also occurs. This may include documentation using point-and-click, computer dictation, or direct provider entry into the EHR. Because the department maintained dictation for the entire study period, the confounders of learning this new system and comfort with the new system of dictation were absent, also limiting generalizability.

This study was conducted at a single academic urban ED with an average ED discharge time well below the national average (12). Also, this study did not analyze many aspects of a complex emergency care system such as patient safety, quality, user satisfaction, patient satisfaction, and differences in system selection. A baseline period of 12 months and comparison period of 12 months were selected to attempt to incorporate the full impact over a 1-year reporting period. Patient volume dropped 3% over the 12 months prior, admission rates dropped 7%, and transfers dropped 29%. It is uncertain what effect this change in volume had on the overall ED metrics. Had volume and acuity level not dropped after EHR implementation, the increase in time metrics could have been even more significant due to ED crowding.

Other confounders included the disposition-to-admit departmental initiative as well as the method of time documentation pre- and post-EHR implementation. Times prior to EHR implementation were based on handwritten times on paper documentation sheets, whereas times in the EHR were obtained from a mouse click in the system. There is a certain amount of unknown variability between the two different methods of collection for the ED time metrics. More studies on these metrics are needed for full understanding of the impact of implementation.

Overall, the study hypothesis was confirmed that an EHR system would have a negative impact on ED metrics at a single institution using a stepwise approach to EHR implementation. Further study is required to find other impacts of mandated EHR implementation and what potential improvements can be made.


1. Shekelle PG, Morton SC, Keeler EB. Costs and benefits of health information technology. Evid Rep Technol Assess. 2006;132:1–71. [PubMed]
2. Handel DA, Hackman JL. Implementing electronic health records in the emergency department. J Emerg Med. 2010;38(2):257–263. [PubMed]
3. Han YY, Carcillo JA, Venkataraman ST, Clark RSB, Watson RS, Nguyen TC, Bayir H, Orr RA. Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2005;116(6):1506–1512. [PubMed]
4. Campbell EM, Sittig DF, Ash JS, Guappone KP, Dykstra RH. Types of unintended consequences related to computerized provider order entry. J Am Med Inform Assoc. 2006;13(5):547–556. [PMC free article] [PubMed]
5. Ward MJ, Froehle CM, Hart KW, Collins SP, Lindsell CJ. Transient and sustained changes in operational performance patient evaluation and medication administration during electronic health record implementation in the emergency department. Ann Emerg Med. 2014;63(3):320–328. [PMC free article] [PubMed]
6. Institute of Medicine. The future of emergency care in the United States health system. Acad Emerg Med. 2006;13(10):1081–1085. [PubMed]
7. Derlet RW, Richards JR. Overcrowding in the nation's emergency departments: complex causes and disturbing effects. Ann Emerg Med. 2000;35(1):63–68. [PubMed]
8. Centers for Disease Control and Prevention. National Hospital Ambulatory Medical Care Survey Factsheet: Emergency Department. August 2011. Available at; accessed January 25, 2017. [PubMed]
9. Washington DC: National Quality Forum, 2008; National Quality Forum endorses measures to address care coordination and efficiency in hospital emergency departments [Press release October 29 2008]
10. The Joint Commission. Patient flow through the emergency department. R3 Report: Requirement, Rational, Reference, 4, December 19, 2012. Available at; accessed January 25, 2017.
11. McHugh M, Van Dyke K, McClelland M, Moss D. Improving Patient Flow and Reducing Emergency Department Crowding: A Guide for Hospitals [AHRQ Publication No. 11(12)-0094]. Rockville, MD: Agency for Healthcare Research and Quality, October 2011
12. Punke H. 46 hospitals with fastest ED discharge time. Becker's Infection Control & Clinical Quality, December 9, 2014. Available at; accessed January 25, 2017.

Articles from Proceedings (Baylor University. Medical Center) are provided here courtesy of Baylor Health Care System