Objective: Implementation of health information technology (HIT) is a national priority to improve patient safety, yet little is known about how electronic charting affects workflow and patient care in busy, fast-paced hospital units. Labor and delivery units are high-risk and high-cost environments in which failures in data transmission or delays in patient care can have tragic consequences. We evaluated the impact of the introduction of an inpatient electronic health record (EHR) on clinical workflow in a high-volume labor and delivery unit in a large university hospital in the United States. Methods: A work-sampling study was performed before and after implementation. Objective observers recorded workflow activities for 3.5-hour periods over nine work shifts (day, evening, night) during 2-week study periods before and after EHR implementation. Activities were standardized to counts per shift and compared using Wilcox two-sample tests. Results: For all health care workers, after introduction of an EHR, direct patient care activities increased from a mean of 12.0 to 15.4 (P = 0.004); computer activities increased from 1.9 to 8.5 (P <0.0001); and personal/idle time decreased from 3.1 to 1.4 (P = 0.0002). Conclusion: The introduction of an EHR into a busy labor and delivery setting did not reduce time spent in direct patient care activities; instead, direct patient care activities increased.

Introduction

In 1999, the Institute of Medicine (IOM) brought the world’s attention to the patient safety vulnerabilities of the U.S. health care system and emphasized the need for widespread adoption of electronic health records (EHRs) as a fundamental component of a new health information technology (HIT) infrastructure designed to improve health care quality.¹ Little research has been done on the impact of HIT, such as EHRs and other interventions, on patient care and safety in obstetrics. Given that childbirth is the leading reason for hospitalization in the United States, comprising over 4 million hospital discharges each year, pregnant women and infants are particularly at risk for safety issues,² making evaluation of the impact of EHRs on obstetric care especially timely.

EHRs have yet to be widely implemented in the United States,³ but data on the impact of these systems on patient safety are conflicting. The use of EHRs with embedded clinical decision support (CDS) can improve adherence to clinical care guidelines,⁴ shorten the length of in-hospital stay,⁵ and improve overall clinical documentation completeness, legibility, and understandability when compared to traditional paper-based medical records.^{6, 7} However, a recent, large study suggested that EHRs are not associated with better quality of care.⁸

Additionally, significant barriers have been identified as limiting ready adoption of these systems. The most commonly cited barriers include high implementation costs, poor integration with legacy systems, fear of technology failure, potential for new kinds of errors, and strong physician resistance due to concerns that practice disruption and loss of clinical productivity are inevitable, regardless of the gains in safety and efficiency the technology might afford.^{3, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20}

Relatively few studies have evaluated EHRs with respect to their impact on clinical work when compared with the larger body of work on the effects of EHRs on physician and/or patient satisfaction, medication error reduction, clinical guideline compliance, risk reduction, and patient outcomes.^{21, 22, 23} Of the studies that have evaluated the impact of EHR implementation on clinical work, the systems under evaluation were found to support both ordering and charting activities, but the studies did not report on time utilization specific to clinical documentation alone.^{24, 25, 26, 27, 28} In addition, although there is a growing body of research on how EHRs impact nursing care activities, very few studies have focused on how EHRs affect the amount of time physicians spend in direct patient care activities.^{21, 25, 29, 30} Finally, we can find no research on the implementation or use of EHRs in the obstetric setting, an area noted to be lagging behind other specialties in EHR adoption.³¹ The aim of the larger patient safety health information technology (HIT) study^a is to systematically evaluate the value of incremental advancements in HIT integration for patient safety and clinical care. This study focuses on the impact of an inpatient electronic obstetric charting system on clinical workflow in a fast-paced, high-volume labor and delivery (L&D) unit.

Methods

HIT Intervention

STORC is a comprehensive obstetric charting system designed with the concurrent goals of facilitating clinical care, enabling clinical outcomes data collection, and promoting patient safety. Incremental advances in systems and data integration were released in series to enable evaluation of value enhancements with each release. In its final, fully integrated form, STORC:

• Integrates existing, disparate data sources into a single point-of-care clinical application (e.g., laboratory results reporting and admission, outpatient and inpatient integration, discharge, and transfer data).
• Pulls key clinical (i.e., pregnancy dating; medical, surgical, and obstetric history; allergies), laboratory, and demographic data collected during prior visits or from hospital systems directly into note fields for editing.
• Provides clinical decision support relevant to obstetrics.
• Provides shortcuts and other tools to speed up care activities (e.g., default values, tailored pick lists, calculators to estimate gestational ages).
• Prints documents in standardized formats.
• Prints patient educational materials and discharge instructions in English or Spanish.
• Keeps clinicians apprised of current clinical studies and patient qualifications for enrollment and more.
• Does not provide clinical order entry functionality.

Figure 1 shows an example of the STORC interface, and Figure 2 provides an example of STORC documentation output for the paper medical record.

Figure 1

Example of STORC interface.

Figure 2

Example of formatted output from STORC.

Study Design

Work-sampling. Work-sampling studies seek to identify the tasks clinicians perform at predetermined, discrete time intervals, so that inferences can be made regarding the overall time a clinician engages in these activities during a given time period. We adapted the work-sampling approach utilized by Fontaine, et al.,³² because this method allows a single researcher to make multiple observations; it works well in clinical settings where staff work is generally restricted to a single physical location (e.g., an inpatient obstetric unit); and it allows researchers to “blend in” more readily in the practice environment, thus reducing the potential for performance bias.³³ Table 1 describes the formal observations and abstractions we used to identify clinical workflow activities.

Table 1

Work activity categories and their operational definitions.

Observations. Observational sessions took place in 2-week blocks on the L&D Unit. A single researcher conducted three, 3.5-hour observation sessions during each of three standard work shifts (i.e., 7 am-3 pm, 3 pm-11 pm, and 11 pm-7 am) for a total of nine observation sessions during each 2-week block. Use of a single researcher eliminated interobserver bias; the observational periods were spread across the three daily shifts to assure collection of a comprehensive and representational set of work tasks.

At the start of each sampling session, the researcher obtained a list of the health care providers (i.e., nurses, medical residents, medical faculty) scheduled to work during that period. Nurse midwives were excluded from this study because they have a lower volume of patient care and would not be expected to be on L&D many times, if at all, during the observation period. All providers on duty were randomly assigned observation times throughout a 3.5-hour observation period.

The researcher cycled through the list, observing the work activities of each provider every 10 minutes, using an obstetric workflow abstraction form to record the observations. When the next provider could not be located in any L&D room, the activity was listed as “off floor,” and the next provider on the list was located for observation. Providers had the option to decline observation; in these cases the observation for that provider for the specific time interval became “declined participation.” When providers were in patient rooms, the provider was assumed to be involved in direct patient care activities and recorded accordingly. When a provider was involved in simultaneous activities, one of which involved direct patient care (e.g., talking with a patient’s family member while on hold on the telephone), the researcher recorded the direct patient care activity as the primary activity.

Results

Basic Study Characteristics

Table 2 describes the basic characteristics of the study, including the number of observations for each type of provider and the unit workload (estimated by the ratio of nurses to patients). A total of 195 observations were obtained over the two study periods: 61.5 percent of observations involved nurses; 31.8 percent involved residents; the remaining 6.7 percent involved medical faculty.

Table 2

Characteristics of observational period.

As shown in Table 3, both the counts of computer-related activities (1.9 vs. 8.5, P <0.0001) and direct patient care (12.0 vs. 15.4, P = 0.004) increased significantly following STORC implementation. Similar patterns were observed for nurses and residents.

Table 3

Working pattern by provider type.

It is also notable that counts of nurses’ activities related to gathering and checking medical records (1.5 vs. 3.0, P = 0.002) increased after STORC EHR implementation. Although comparisons for faculty were not statistically significant due to small sample size, activity counts for computer work (0.3 vs. 4.2), direct patient care (7.8 vs. 8.8), and talking to nurses or residents (8.2 vs. 11.4) all increased after the implementation of STORC. Because talking to other workers is a vital component of direct patient care, we grouped direct patient care activities and talking with other workers together under the header “patient-related work” to more fully assess the impact of STORC on these activities. Patient-related work activities increased significantly (20.1 vs. 23.9, P = 0.001) overall, with residents having the greatest activity count increase (21.3 vs. 25.8, P = 0.005), followed by nurses (13.0 vs. 16.1, P = 0.02).

Although the amount of paperwork did not seem to decrease (4.3 vs. 4.5 counts per shift), personal/idle waiting time decreased from 4.1 to 1.8 counts per shift. Activity counts for other recorded activities—such as telephone/fax use, recording, environmental maintenance, and traveling—were very low before STORC implementation, and no significant difference was detected (data not shown).

Direct Patient Care Activities

Comparisons of direct patient care activity counts before and after STORC implementation are summarized in Figure 3. Even after adjusting for workload, direct patient care activity counts showed a statistically significant increase for nurses (13.0 vs. 16.1, P = 0.04) and residents (10.9 vs. 15.4, P = 0.02). Although activity counts for attending staff increased, these differences were not significant. Overall, direct patient care activity increased significantly (P = 0.03) following implementation of STORC.

Figure 3

Direct patient care before and after STORC implementation. (*P <0.05).

Discussion

Our results suggest that an EHR can be successfully implemented in busy, fast-paced, procedure-oriented hospital units without negatively affecting activities directly involving patients. We believe this finding is very important to patients, providers, hospitals, and policymakers, particularly during childbirth, when fetal status can change in minutes, but the experience leaves a permanent memory for families.

We expected computer work to increase with the introduction of an EHR because the system marked the formal shift from pen-and-paper documentation to the computer. However, we were pleased to discover that this increase did not appear to come at the expense of direct patient care work. In fact, direct patient care activity counts significantly increased for nurses and medical residents, suggesting that EHRs, like STORC, might improve practice efficiency in other areas, despite the greater time spent at the computer. For example, direct importing of laboratory, prenatal visit, and scheduling data into the electronic patient record might have reduced the amount of time clinicians spent locating and collating this information from disparate sources in order to compile and synthesize sufficient data to provide care. Additionally, embedded calculators for determining due dates (estimated dates of conception), Bishop’s scores, and preconfigured selections (e.g., pick lists, menus) may have led to time savings for clinicians.

Finally, the amount of time spent repeatedly transcribing these data points from one form to another may have been significantly reduced. This is because STORC is designed to pull data forward as collected, so that once the information is recorded, it populates all forms where this information is required (while allowing editing).

The increase in activity counts for gathering and checking medical records might have resulted from STORC’s lack of order-entry functionality. This means that clinical orders were still written into the paper medical record, requiring providers to locate the paper record. In addition, this EHR had not implemented the unit “white board” electronically. The white board was updated frequently to reflect the most up-to-date, at-a-glance information for all patients in the unit, and thus required constant management.

The EHR implementation also coincided with the introduction of a new maternal/fetal monitoring system on the L&D unit. The new work associated with this system (e.g., reading monitoring strips, documenting interventions) is most likely reflected in the increased time spent processing these new data and interacting with the monitors.

The reduction in idle waiting activity counts likely resulted from the improved availability of patient information in a single electronic source. For example, prior to STORC implementation, when a woman arrived for an unscheduled delivery, the L&D staff spent significant time locating the paper prenatal record or contacting the ambulatory clinics to have copies of the record faxed to the unit. Because this information is now collected electronically, it is available immediately when the woman arrives for delivery, eliminating this often frustrating search for important historical clinical data.

Overall, the shifts in types and amounts of work activities were reasonable with the introduction of the computerization of clinical documentation. The time-saving improvements with the technology (e.g., single source of information, prenatal visit information available at the time of delivery) did not entirely offset the increased time required to document patient care. However, these changes did not appear to negatively affect total direct patient care activities, despite concerns to the contrary.

We believe this study provides an important view of the positive value that HIT interventions can have on clinical care for high-reliability units if they facilitate integration of data across systems, saving clinicians time and ultimately improving patient care and safety.

Conclusion

The introduction of a clinical information system into a busy L&D setting did not reduce the total count of direct patient care activities. This study may assuage physician fears about the potential for loss of direct patient care time due to documentation time spent on electronic systems. Although overall computer work increased, this was not to the detriment of patient care. The increase in computer work is an unavoidable by-product of the technology age. This is not to say that the shift from paper to computer is seamless, effortless, or easy. The shift does require that clinicians rework their routines, which alone can cause strong emotional reactions and resistance to change. Happily, we see an overall increase in patient-related work, which we believe translates directly to higher quality care in the obstetric setting.

Acknowledgment

This research was supported by the Agency for Healthcare Research and Quality (AHRQ Grant No. R01 HS015321).

References

1.

Institute of Medicine, Committee on Quality Health Care in America. To err is human: Building a safer health system. Washington, DC: National Academies Press; 1999.

2.

DeFrances C, Podgornik M. Advance data from vital and health statistics. Hyattsville, MD: National Center for Health Statistics; 2005. National Hospital Discharge Survey.

3.

Ash JS, Gorman PN, Seshadri V, et al. Computerized physician order entry in U.S. hospitals: Results of a 2002 survey. J Am Med Inform Assoc. 2004;11:95–99. [PMC free article: PMC353025] [PubMed: 14633935]

4.

Balas E, Weingarten S, Garb C, et al. Improving preventive care by prompting physicians. Arch Intern Med. 2000;160:301–308. [PubMed: 10668831]

5.

Kuperman GJ, Gibson RF. Computer physician order entry: Benefits, costs, and issues. Ann Intern Med. 2003;139:31–39. [PubMed: 12834316]

6.

Van der Meijden MJ, Tange HJ, Troost J, et al. Determinants of success of inpatient clinical information systems: A literature review. J Am Med Inform Assoc. 2003;10:235–243. [PMC free article: PMC342046] [PubMed: 12626373]

7.

Hippisley-Cox J, Pringle M, Cater R, et al. The electronic patient record in primary care – regression or progression? A cross-sectional study. Br Med J. 2003;326:1439–1443. [PMC free article: PMC162256] [PubMed: 12829558]

8.

Linder J, Ma J, Bates D, et al. Electronic health record use and the quality of ambulatory care in the United States. Arch Intern Med. 2007;167:1400–1405. [PubMed: 17620534]

9.

Ash J, Bates D. Factors and forces affecting EHR system adoption: Report of a 2004 ACMI discussion. J Am Med Inform Assoc. 2005;12:8–12. [PMC free article: PMC543830] [PubMed: 15492027]

10.

Han Y, Carcillo J, Venkataraman S, et al. Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2005;116:1506–1512. [PubMed: 16322178]

11.

Koppel R. Computerized physician order entry systems: The right prescription? LDI Issue Brief. 2005;10:1–4. [PubMed: 15810170]

12.

Loomis GA, Ries JS, Saywell RM, et al. If electronic medical records are so great, why aren't family physicians using them? J Fam Pract. 2002;57:636–641. [PubMed: 12160503]

13.

Massaro TA. Introducing physician order entry at a major academic medical center: I. Impact on organizational culture and behavior. Acad Med. 1993;68:20–25. [PubMed: 8447887]

14.

Massaro TA. Introducing physician order entry at a major academic medical center: II. Impact on medical education. Acad Med. 1993;68:25–30. [PubMed: 8447888]

15.

Miller R, Sim I. Physicians' use of electronic medical records: Barriers and solutions. Health Aff. 2004;23:116–126. [PubMed: 15046136]

16.

Oren E, Shaffer ER, Guglielmo BJ. Impact of emerging technologies on medication errors and adverse drug events. Am J Hosp Pharm. 2003;60:1447–1458. [PubMed: 12892029]

17.

Poissant L, Pereira J, Tamblyn R, et al. The impact of electronic health records on time efficiency of physicians and nurses: A systematic review. J Am Med Inform Assoc. 2005;12:505–516. [PMC free article: PMC1205599] [PubMed: 15905487]

18.

Sprague L. Electronic health records: How close? How far to go. NHPF Issue Brief. 2004:1–17. [PubMed: 15470818]

19.

The Leapfrog Group. Factsheet: Computer physician order entry. [Accessed March 11, 2008]. Available at: www​.leapfroggroup.org​/for_hospitals/leapfrog​_safety_practices/cpoe.

20.

Wears RL, Berg M. Computer technology and clinical work: Still waiting for Godot. JAMA. 2005;293:1261–1263. [PubMed: 15755949]

21.

Delpierre C, Cuzin L, Fillaux J, et al. A systematic review of computer-based patient record systems and quality of care: More randomized clinical trials or a broader approach? Int J Qual Health Care. 2004;16:407–416. [PubMed: 15375102]

22.

Johnston M, Langton K, Haynes R, et al. Effects of computer-based clinical decision support systems on clinician performance and patient outcome. A critical appraisal of research. Ann Intern Med. 1994;120:135–142. [PubMed: 8256973]

23.

Shiffman R, Liaw Y, Brandt C, et al. Computer-based guideline implementation systems: A systematic review of functionality and effectiveness. J Am Med Inform Assoc. 1999;6:104–114. [PMC free article: PMC61349] [PubMed: 10094063]

24.

Bates DW, Boyle DL, Teich JM. Impact of computerized physician order entry on physician time. Proc Annu Symp Comput Appl Med Care. 1994 [PMC free article: PMC2247898] [PubMed: 7950101]

25.

Overhage JM, Perkins S, Tierney WM, et al. Controlled trial of direct physician order entry: Effects on physicians' time utilization in ambulatory primary care internal medicine practices. J Am Med Inform Assoc. 2001;8:361–371. [PMC free article: PMC130081] [PubMed: 11418543]

26.

Pizziferri L, Kittler AF, Volk LA, et al. Primary care physician time utilization before and after implementation of an electronic health record: A time-motion study. J Biomed Inform. 2005;38:176–188. [PubMed: 15896691]

27.

Shu K, Boyle D, Spurr C, et al. Comparison of time spent writing orders on paper with computerized physician order entry. Medinfo. 2001;10:1207–1211. [PubMed: 11604922]

28.

Tierney WM, Miller ME, Overhage JM, et al. Physician inpatient order writing on microcomputer workstations. Effects on resource utilization. JAMA. 1993;269:379–383. [PubMed: 8418345]

29.

Apkon M, Singhaviranon P. Impact of an electronic information system on physician workflow and data collection in the intensive care unit. Intensive Care Med. 2001;27:122–130. [PubMed: 11280623]

30.

Cordero L, Kuehn L, Kumar RR, et al. Impact of computerized physician order entry on clinical practice in a newborn intensive care unit. J Perinatol. 2004;24:88–93. [PubMed: 14872207]

31.

Menachemi N, Lee SC, Shepherd JE, et al. Proliferation of electronic health records among obstetrician-gynecologists. Qual Manag Health Care. 2006;15:150–156. [PubMed: 16849986]

32.

Fontaine BR, Speedie S, Abelson D, et al. A work-sampling tool to measure the effect of electronic medical record implementation on health care workers. J Ambul Care Manage. 2000;23:71–85. [PubMed: 11184897]

33.

Finkler SA, Knickman JR, Hendrickson G, et al. A comparison of work-sampling and time-and-motion techniques for studies in health services research. Health Serv Res. 1993;28:577–597. [PMC free article: PMC1069965] [PubMed: 8270422]