This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Epstein, R. H. Articles by Piotrowski, E. Search for Related Content PubMed PubMed Citation Articles by Epstein, R. H. Articles by Piotrowski, E. Related Collections Economics and Health Care Research

---

ECONOMICS, EDUCATION, AND POLICY

Automated Correction of Room Location Errors in Anesthesia Information Management Systems

Richard H. Epstein, MD^*, Franklin Dexter, MD, PhD, and Elizabeth Piotrowski, RN, MA

From the *Department of Anesthesiology, Jefferson Medical College, Philadelphia, Pennsylvania; Departments of Anesthesia and Health Management and Policy, University of Iowa, Iowa City, Iowa; and Department of Nursing, Thomas Jefferson University Hospital.

Address correspondence and reprint requests to Richard H. Epstein, MD, Thomas Jefferson University Hospital, 111 S. 11th St. Suite 5480G, Philadelphia, PA 19107. Address e-mail to richard.epstein{at}jefferson.edu.

Abstract

BACKGROUND: Anesthesia information management systems (AIMS) and operating room information management systems (ORIMS) are both used in operating rooms (OR). Anesthesia providers use AIMS to document their care in near real-time, including milestone events, and these systems automatically record vital signs from patient monitors. Circulating nurses use ORIMS primarily to document procedural information. Because of automatic documentation, AIMS would be ideal platforms for OR managerial decision support if the correct locations of cases in progress were known accurately. Trust is diminished if recommendations are poor.

METHODS: We compiled room location error rates from prior analyses of ORIMS data. Data from 24 consecutive 4-wk periods (45,459 cases) were analyzed from one hospital where both ORIMS and AIMS data were available. The actual location of cases was inferred from the physical location of the workstation recording the majority of pulse oximetry saturations. These were compared to the listed location in the AIMS and the final corrected location in the ORIMS. The scheduled and final ORIMS locations were compared to determine how often location changes were updated before the start of anesthesia. The location of cases was inferred in near real-time by using the identifier of the AIMS workstation transmitting pulse oximetry saturated electrocardiogram heart rate, and end-tidal CO₂ partial pressures.

RESULTS: Location error rates ranged from 0% to 7.5% at 42 hospitals. The error rate at the studied hospital was just 0.4%, showing that the hospital was suitable for investigation. The 0.4% error rate was based on cases listed as overlapping in the same OR, and thus under-estimated the actual error rate in the ORIMS (1.0%). With education, there was a decrease in the moved cases in the ORIMS whose location was not changed before the start of anesthesia (9.3%–2.0%, P < 10^–5). Despite the significant improvement (P < 10^–5) in the error rate between the AIMS listed and actual locations, the residual AIMS real-time error rate was 4.1% of cases. Use of vital sign data reduced errors to <0.1%.

CONCLUSIONS: Education can only modestly improve the accuracy of OR locations in ORIMS and AIMS data. The actual location can be inferred, either in near real-time or afterwards, from the AIMS workstation transmitting vital sign data. This addresses the fundamental problem of cases having more than one location during the course of anesthetic care (e.g., holding area, block room, OR, and postanesthesia care unit), which cannot be determined from scheduled ORIMS or listed AIMS locations.

Anesthesia information management systems (AIMS) and operating room information management systems (ORIMS) are both used in operating rooms (ORs). Anesthesia providers use AIMS to document their care in near real-time, including milestone events (e.g., induction, tracheal extubation), and these systems automatically record vital signs from patient monitors. Circulating nurses use ORIMS primarily to document procedural information, although there is some duplication in the events recorded in the two systems. AIMS are used increasingly for managerial purposes, including enhancing billing accuracy, reducing drug costs, and accurately matching staffing to workload.1–10 AIMS are a more rational data source for management decision support on the day of surgery than ORIMS, because data collected automatically from patient monitors can be used to infer room occupancy,11,12 and the clinical events documented by anesthesia care providers are useful for judging the time remaining in cases (e.g., "closing" or "neostigmine 2.5 mg IV"). Furthermore, AIMS are used in non-OR locations where anesthesia care is increasingly being provided (e.g., block rooms, endoscopy suites, and radiology), whereas ORIMS are restricted to ORs.

A recent study of simple OR management scenarios demonstrated that decisions made using status displays of ongoing OR cases can be correct at a rate no better than by chance alone.13 In contrast, specific recommendations improved performance.13 Thus, decision support system(s), as compared to electronic whiteboards only displaying data about each case, likely will be needed to improve OR management on the day of surgery.13 An important challenge is that ORIMS and AIMS contain errors in the locations where cases were performed.14,15 Although errors in location do not substantively affect long-term OR management decisions, such as staffing and case scheduling,14,15 decision-support recommendations on the day of surgery require knowing the locations of anesthetics that are in progress.13,16 Such decisions include moving cases, deciding where to schedule add-on cases, determining which case to start next, and assigning anesthesia providers.16

Whether people will follow automated OR management recommendations will depend on whether they trust the reliability of the recommendations.13 This issue of trust related to automated decision support systems has been analyzed extensively outside the field of anesthesiology (e.g., aviation and military weapons targeting, as listed in Table 5 of Ref. 13). To summarize briefly, automation is used when a user’s trust in the recommendation exceeds the user’s self-confidence, and manual control is used when the opposite is true.17 Unsafe recommendations (e.g., to send an inappropriate anesthesia call team to relieve the day staff) are recognized by anesthesia providers and reduce trust in recommendations.13

We examined OR location error rates in datasets from a cross-section of hospitals, and studied in detail the error rates in the ORIMS and AIMS databases at a hospital with both systems in place. We describe a simple method to nearly eliminate such errors at facilities with an AIMS.

METHODS

Data Collection
After a determination by its IRB that "the study does not qualify as human subjects research" the primary data analyzed for this study were retrospective, permanently de-identified data from Hospital A.

As a preliminary step, we compared the error rates of Hospital A’s listed case locations with the error rates from 41 other hospitals, representing a mix of academic and private practice facilities. We hypothesized that Hospital A had a relatively low error rate, making the hospital suitable for the primary analyses. The error rate from Hospital B was provided for reference purposes (Fig. 1), because multiple prior OR management studies16 demonstrated that it had a highly accurate database. The error rates for the 41 other hospitals were obtained from assessments of specialty-specific staffing.14 Before appropriate staffing was calculated based on reducing expected under-utilized and over-utilized OR time,14 a check was made for multiple cases being listed in the same OR at the same time.15 For each pair of overlapping cases, one location was considered arbitrarily to be correct, and the other, incorrect.15 The error rate was calculated as the number of cases with an incorrect location divided by the total number of cases. For an error in location to be detected based on the overlapping of cases in the same location, the duration of overlap had to exceed the turnover time between cases. At most hospitals, this meant that the duration of overlap was longer than 30 min.18 Thus, these errors in location do not represent trivial overlaps of a few minutes from use of multiple unsynchronized clocks showing different times.

View larger version (25K): [in this window] [in a new window]   Figure 1. Room error rates in the finalized operating room (OR) datasets from 42 hospitals. The hospitals include both academic and non-academic facilities. A case location was considered to be in error if there was another case listed in the same room for some portion of the case (time in the OR until time out of the OR),14,15 Thus, one of the two cases was counted as correct, and the other incorrect. These are not trivial overlaps of a few minutes, possibly attributable to the use of clocks with different times, but encompass the turnovers between cases, often more than 30 minutes.18 The error rate was calculated by dividing the number of cases with incorrect locations divided by the total number of cases in the analyzed dataset. The error rate from Hospital B is shown, because those data have been used for many previously published studies of case duration prediction.16

The following detailed analyses were performed on the data from Hospital A. The location for each case was determined from the "room location" field of the AIMS (Innovian, Dräger Medical, Inc, Telford, PA). We also recorded the "scheduled location" and "final location" fields from the ORIMS (ORSOS, formerly Per-Sé Technologies, Alpharetta, GA).

At Hospital A, updates to the scheduled location in the ORIMS are sent automatically to the AIMS via a Health Level 7® (HL7) interface until the case is opened in the AIMS, at which time further updates are not processed by the AIMS. Updates to the scheduled location by the ORIMS are also blocked once the case is opened in the ORIMS on the OR nursing computer. There are two dedicated full time equivalent OR secretaries who make schedule changes on the day of surgery from 7 am to 3 pm, and one secretary who performs this function from 3 pm until 11 pm. The OR charge nurse makes these changes from 11 pm until 7 am. In addition, there are full time OR schedulers who book new surgical cases and participate in a post hoc data cleaning. Both the anesthesia care provider and the circulating nurse taking care of the patient are also supposed to check the cases they have open in their respective systems and correct the OR location if the case was moved but not updated in the ORIMS by the OR secretaries. In the AIMS, the listed location field is overwritten, with the change posted to an audit table. In the ORIMS, the new location is written to the final location field, but the scheduled location field is left as it was when the case was opened in the ORIMS on the OR nursing computer.

The final location in the ORIMS is audited the next business day via an automated report that identifies overlapping cases, as described above for Figure 1. Cases with temporary "locations" (e.g., "first available room," "emergent") are also flagged. The final location is corrected by referencing handwritten annotations on the master OR schedule maintained by the OR secretaries and/or by matching the anesthesia providers on the OR schedule to the anesthesia providers listed for the case in the ORIMS.

We compared the ORIMS scheduled location to the final location to measure corrections to location made after the anesthetic started. The data analyzed were the 45,459 cases scheduled and performed in the ORIMS from November 6, 2005 to September 6, 2007. The starting date was chosen because it was the first Sunday two full weeks after implementation of the AIMS, giving anesthesia providers some time to be comfortable with the software. The ORIMS had already been in use for several years. The final date was chosen based on its providing 24 consecutive 4-wk periods of data for analysis.14

The method we developed to identify the actual location for each case in near real-time relies on the identifier of the AIMS workstation that is transmitting pulse oximetry saturation, electrocardiogram derived heart rate data, or the end-tidal CO₂ partial pressure, and is based on the algorithm in the Appendix of Ref. 11. Since all patients receiving anesthetic care at Hospital A are monitored with electrocardiogram and oximetry monitors, and capnometry is used during most anesthetics, the incidence of apparent gaps during cases is mitigated (i.e., temporary disconnections or signal acquisition problems from all three sources, simultaneously, is very rare).11,12 Each AIMS computer is permanently mounted in every anesthetizing location and has a workstation identifier field in the database linked to another field identifying the physical location of the AIMS computer. A case must be opened on the computer and trending initiated manually by the provider for data to be transmitted. Thus, even if the case starts in one location and ends in another location, one automatically knows the location where the case is currently taking place.

For example, at Hospital A, if a patient scheduled for hip replacement surgery in OR 1 is taken to a block room for administration of a spinal anesthetic, the AIMS case is initiated in that block room. When the patient is transported to OR 1, the AIMS case is closed in the block room and, upon arrival in OR 1, the case is reopened. The workstation identifier defines a location for each case in near real-time.

We defined the actual location as the location of the AIMS workstation in an OR or block room that sends the most pulse oximeter signals during the anesthetic. Because the great majority of each anesthetic takes place in the room where the surgical or diagnostic procedure is performed, the most common source of the pulse oximetry saturation signal indicates the actual location. This is especially important at Hospital A, because most hip and knee replacement surgeries (approximately 2400 per year) are started in a block room under regional anesthesia, and thus have two separate locations during the case. We could not reliably infer actual locations from the workstations where defined events were entered, because the anesthetic record was sometimes modified after the fact from a different workstation (e.g., the time of incision was modified after the case ended) or the event was missing altogether.

To assess the accuracy of the AIMS location, we compared the final value (in the database) of the AIMS room location to the actual location. Because the anesthesia provider can correct the AIMS room location during or after the case, this approach will under-estimate the rate of real-time location errors. The 59,778 cases studied (Fig. 2) excluded anesthetics performed in mobile locations (e.g., magnetic resonance imaging and interventional radiology), because our mobile workstations are identified with their function (e.g., "Mobile 1") rather than a specific location. The frequency of errors was calculated pairwise for each of the 24 4-wk periods. The statistical analysis for the change over time (e.g., with education) was based on the first and last 6 4-wk periods.

View larger version (17K): [in this window] [in a new window]   Figure 2. Room error rates in the listed location (•) in the anesthesia information management system (AIMS) and the final corrected location (x) in operating room information management system (ORIMS) were calculated against the actual case location rates as determined from identifying in the database the location of the AIMS workstation that recorded the most pulse oximetry saturations throughout the anesthetic. Data were binned in 24 consecutive 4-wk epochs. Consult the Methods section for details of the analysis procedure. The apparent spike in error rates during the 10th through 12th 4-wk period was most likely related to a reduction in the number of OR schedulers from five to two due to attrition and the vacation of the OR manager during the 11th 4-wk period. The full complement of five OR schedulers was restored in the 13th period, The AIMS error rate follows the ORIMS error rate because the initial AIMS locations are populated with the ORIMS scheduled locations. Overall, there was no improvement over time (P = 0.39) in the ORIMS final location rate. However, the AIMS location error rate did improve over time (P < 10–5) from an initial value of 9.0% (95% confidence interval CI 8.1%–9.9%) to 4.1%, (95% CI 3.7%–4.5%) for the last 6 4-wk periods. Thus, attempts to use the listed OR location in the AIMS to determine where active cases are located will be subject to an error of at least 4.1%, since corrections to the listed AIMS location were made at various times during the course of the anesthetic. As described in the Results, if the workstation identifier sending the final intraoperative event were used instead of the method based on vital signs, then the error rates would be an additional 0.4% higher.

We compared the ORIMS final location with the actual location (Fig. 2). There were 40,231 cases studied for this comparison, based on the following exclusions. Comparison was limited to cases performed at locations where both the ORIMS was used intraoperatively (e.g., not the endoscopy suite) and where anesthesia providers used the records created by the transmission of the scheduled case from the ORIMS.

Finally, our approach to determining the actual location of a case depends on accuracy of the location linked to the AIMS workstation identifier. We query the database every 15 min, and if the same OR location is assigned to more than one workstation or if the location is unassigned, a text message is sent to the cell phone of the AIMS clinical administrator on call. All such occurrences and the time needed to correct the problem were documented. They are reported qualitatively, below.

Education Effect
The effect of training on improving the accuracy of the room information data in the AIMS was assessed in a natural experiment. The initial AIMS implementation had an unacceptably high rate of inaccurate room location data, which was attributed to transmission of an inaccurate scheduled location field from the ORIMS system to the AIMS system at the beginning of the anesthetic. A concerted effort was made to encourage the OR scheduling staff to change scheduled locations in the ORIMS before the start of anesthesia. These efforts included announcements from OR managers, emailed reminders, meetings dedicated to the problem, documentation of individual and group error rates, and counseling statements in individual performance files. Anesthesia providers and circulating nurses also received repeated reminders to correct any OR locations that were not updated in their respective systems. This effort afforded us an opportunity to examine the effects of education and training on improving the accuracy of room location data in the AIMS systems.

Statistical Analysis
The data were binned by 4-wk period to eliminate variations by time of day and by day of week. The numbers of cases with and without a different scheduled vs final ORIMS location during each 4-wk period were transformed using the Freeman-Tukey double arcsin transformation, as explained in Appendix 2 of Ref. 19. Student’s two-sided t-test and t-distribution were used to compare the first and last 6 4-wk periods and to compute corresponding confidence intervals, respectively. The same method was used for the other comparisons.

RESULTS

In the finalized datasets sent by hospitals for assessment of their specialty-specific staffing,14 case location error rates based on overlapping cases in their ORIMS data ranged from 0% to 7.5% (n = 42 hospitals) (Fig. 1). The observed error rate at Hospital A in its ORIMS data was 0.4%, showing that our study was performed at a facility with an error rate lower than that of most (83%) hospitals and close to the error rate of Hospital B (<0.1%). These observed error rates (Fig. 1) under-estimate the true error rates, because they only include misspecifications of locations resulting in an overlap of cases in the same location.15 The actual ORIMS final location error rate at Hospital A was 1.0% (Fig. 2) (95% confidence interval 0.8% to 1.3%).

The AIMS initially listed an incorrect room location in 9% (95% confidence interval 8.1% to 9.9%) of cases (Fig. 2). This was mostly a result of an incorrect room location in the ORIMS that was transmitted to the AIMS. More than 80% of these errors occurred on workdays between 7 am and 5 pm (Fig. 3) (i.e., the errors are not due to inadequate supervision, absence of trained schedulers, and/or too few OR managers available).

View larger version (18K): [in this window] [in a new window]   Figure 3. Percentage of total room errors in the anesthesia information management system during three consecutive 32 wk periods reported according to the shift when the mistakes occurred. n = 1690 for the first, n = 1234 for the second, and n = 778 for the third set of 8 4-wk periods. The primary source of the errors was the failure of the operating room (OR) secretaries to update the case location in the OR information management system before the start of the anesthetic. These errors occurred mostly (>80% of the time) during the 7 am to 5 pm shift, when there were both OR secretaries and OR managers present. Thus, errors were not due to the lack of available personnel or managerial oversight, as might have been the case if most errors occurred at night and on weekends. In the figure, the weekends are from Saturday 7 am to Monday 7 am.

The educational program with quantitative goals and strong supervisory feedback resulted in a significant reduction in the scheduled location at the start of the anesthetic differing from the final location in the ORIMS (Fig. 4) (9.3% to 2.0%, P < 10^–5). Thus, the room location of the case was more frequently being updated in the ORIMS before the anesthetic began. This resulted in a reduction in the error rate of the AIMS room location to 4.1% (P < 10^–5, 95% confidence interval 3.7%–4.5%) (Fig. 2).

View larger version (17K): [in this window] [in a new window]   Figure 4. Difference between the initial location of a case in the anesthesia information management system (AIMS) and the final listed located of the case in the operating room information management system (ORIMS). The data are from Hospital A and were binned in consecutive 4-wk epochs between November 6, 2005 and September 6, 2007. The initial location (sent via an interface from the ORIMS scheduled location to the AIMS) was compared to the final location recorded in the ORIMS, divided by the total number of cases. The final location is a function of corrections made in the OR by the circulating nurses and a post hoc cleaning process performed the next business day after surgery. Consult the Methods section for details of the analysis procedure. There was improvement (P < 10–5) in the initial 9.3% (95% CI 8.2%–10%) rate of change in location resulting from an intensive effort to get the OR schedulers to promptly adjust case locations when cases were moved. However, at the end of the study interval, 2.0% (95% CI 1.4%–2.8%) of OR assignments in the scheduling system were still being changed after the anesthetic started. The consequence of this residual error in location is that the initial location in the AIMS will be incorrect.

The final location differed from the actual location in 0.4% percent of AIMS cases (95% confidence interval 0.3%–0.5%, P < 10^–5). This mostly resulted from users reopening the record at a different workstation than where the case took place and thus unintentionally/unknowingly modifying the final location.

Over the 5 mo that the actual location was determined from the workstation identifier, there were only two cases (0.02%) in which it was inaccurate. Both errors were caused by a replacement workstation installed with an incorrect location. There were also two cases (0.02%) in which a workstation was installed without a location being specified, resulting in the display of the case’s location being "unknown."

DISCUSSION

The ubiquitous problem of incorrect OR locations in ORIMS (Fig. 1) and AIMS data does not affect the accuracy of appropriate calculations for staffing and case scheduling,14,15 but does reduce the accuracy of recommendations on the day of surgery16 and adversely affects trust in those decision support recommendations (see Introduction).13 Errors in locations in both ORIMS and AIMS likely cannot be improved sufficiently through educational processes alone to a level of accuracy required for day of surgery management decision support systems (Figs. 2–4). However, the location where the anesthetic is taking place can be accurately inferred, in near real-time, automatically by the identifier of the AIMS workstation transmitting vital sign data.11,12 While the surgical procedure takes place in one OR, and hence has a single location, this is not necessarily true for the anesthetic, which may start in the holding area, continue in a block room, move to an OR, and finally conclude in a postanesthesia care unit. For the purpose of decision support on the day of surgery, it is necessary to know the location of the patient in near real-time, which cannot be determined from scheduled locations in either the ORIMS or the AIMS. For example, at Hospital A, only one patient could be located in a given OR at any point in time, but there were often two patients scheduled for surgery in the same OR who were simultaneously undergoing anesthesia care by separate providers, one in the block room, and one in the OR. We strongly emphasize the importance of this issue because of its impact on the implementation of AIMS and OR decision support systems.

As an alternate methodology to our use of AIMS, radiofrequency identification and active infrared radiofrequency identification can be used to match patients to specific OR locations. Some of these systems have been configured to send a page to the attending responsible for the case in the event that there is a mismatch with the OR schedule.20,21 However, such systems will generate false pages when the ORIMS scheduled case locations are inaccurate because the detector location is used to find the cases scheduled for that OR.21

Our method reduces the time required for an OR scheduler to fix the ORIMS final location after the fact (see Methods). At Hospital A, the time required to clean one day’s cases ranged from 1 hr for experienced schedulers to 3 hr for inexperienced ones. Although this process does not involve management decision making, the ORIMS database is the official log of OR cases, and thus accuracy is desirable. Our finding that education, goal setting, and retrospective feedback were insufficient to result in the ORIMS scheduled location being updated before the anesthetic began (Fig. 4) was typical,22 not an aberrancy at Hospital A.

Our method works best for locations with permanent AIMS workstations (e.g., ORs and block rooms). If a portable workstation is wheeled back and forth between two rooms, it is unlikely that users will remember to reset the location with any greater accuracy than they update the scheduled case location. For portable AIMS workstations used at multiple locations outside the OR suite (e.g., computerized tomography and magnetic resonance imaging), the method will not provide accuracy, since these workstations are not assigned to a specific location.

Our approach of inferring case location using the physical AIMS workstation transmitting vital sign data to the AIMS database should be of general utility, as the workstation transmitting vital signs is recorded by other systems (e.g., PICIS, GE Centricity Anesthesia, CompuRecord, and DocuSys [personal communications Michael Vigoda, Michael O’Reilly, Sarah Nabel, and Stanley Muravchik, respectively]). Facilities without an AIMS, but which have a networked system to continuously record data transmitted from patient monitors, can infer room occupancy, but not the specific case, from the use of several vital signs.11,12 For processes that require accurate knowledge of where cases are being performed in near real-time and reports that are sensitive to room errors, either method should provide better performance than those relying on human intervention to modify incorrectly specified locations. The system is also easy to troubleshoot. In all four circumstances in which inaccurate terminal location information was present in the system, automatic notification was received by the AIMS administrator and the problem corrected within 0.5 hr.

Lastly, our method was popular with the clinicians caring for surgical patients. The initial educational efforts essentially harassed the anesthesia providers to ensure accurate entry of the room location in the AIMS system. Accurate automatic identification of the room permitted the clinicians to focus on patient care instead.

Footnotes

Accepted for publication April 23, 2008.

Franklin Dexter is editor of Economics, Education, and Policy for the Journal. This manuscript was handled by Steve Shafer, Editor-in-Chief and Franklin Dexter was not involved in any way with the editorial process or decision.

Conflict of Interest: FD is the Director of the Division of Management Consulting of the Department of Anesthesia of the University of IA. He receives no funds personally other than his salary from the State of Iowa, including no travel expenses or honoraria, and has tenure with no incentive program.

REFERENCES

1. Kheterpal S, Gupta R, Blum JM, Tremper KK, O’Reilly M, Kazanjian PE. Electronic reminders improve procedure documentation compliance and professional fee reimbursement. Anesth Analg 2007;104:592–7[Abstract/Free Full Text]
2. Spring SF, Sandberg WS, Anupama S, Walsh JL, Driscoll WD, Raines DE. Automated documentation error detection and notification improves anesthesia billing performance. Anesthesiology 2007;106:157–63[Web of Science][Medline]
3. Vigoda MM, Lubarsky DA. The medicolegal importance of enhancing timeliness of documentation when using an anesthesia information system and the response to automated feedback in an academic practice. Anesth Analg 2006;103:131–6[Abstract/Free Full Text]
4. Sandberg WS, Sandberg EH, Seim AR, Anupama S, Ehrenfeld JM, Spring SF, Walsh L. Real-time checking of electronic anesthesia records for documentation errors and automatically text messaging clinicians improves quality of documentation. 2008;106:192–201
5. Wax DB, Beilin Y, Levin M, Chadha N, Krol M, Reich DL. The effect of an interactive visual reminder in an anesthesia information management system on timeliness of prophylactic antibiotic administration. Anesth Analg 2007;104:1462–66[Abstract/Free Full Text]
6. O’Reilly M, Talsma A, VanRiper S, Kheterpal S, Burney R. An anesthesia information system designed to provide physician-specific feedback improves timely administration of prophylactic antibiotics. Anesth Analg 2006;103:908–12[Abstract/Free Full Text]
7. Lubarsky DA, Sanderson IC, Gilbert WC, King KP, Ginsberg B, Dear GL, Coleman RL, Pafford TD, Reves JG. Using an anesthesia information management system as a cost containment tool. Description and validation. Anesthesiology 1997;86:1161–9[Web of Science][Medline]
8. Dexter F, Epstein RH. Optimizing second shift OR staffing. AORN J 2003;77:825–30[Medline]
9. Reich DL, Kahn RA, Wax D, Palvia T, Galati M, Krol M. Development of a module for point-of-care charge capture and submission using an anesthesia information management system. Anesthesiology 2006;105:179–86[Web of Science][Medline]
10. Okamura A, Murayama H, Sato N, Kemmotsu O. Automated anesthesia billing by electronic anesthesia record keeping system. Masui 1999;48:903–8[Medline]
11. Xiao Y, Hu P, Hao H, Ho D, Dexter F, Mackenzie CF, Seagull FJ, Dutton R. Algorithm for processing vital sign monitoring data to remotely identify operating room occupancy in real-time. Anesth Analg 2005;101:823–9[Abstract/Free Full Text]
12. Xiao Y, Dexter F, Hu P, Dutton RP. Usage of distributed video when real-time occupancy status was available. Anesth Analg 2008;106:554–60[Abstract/Free Full Text]
13. Dexter F, Willemsen-Dunlap A, Lee JD. Operating room managerial decision-making on the day of surgery with and without computer recommendations and status displays. Anesth Analg 2007;105:419–29[Abstract/Free Full Text]
14. McIntosh C, Dexter F, Epstein RH. Impact of service-specific staffing, case scheduling, turnovers, and first-case starts on anesthesia group and operating room productivity: tutorial using data from an Australian hospital. Anesth Analg 2006;103:1499–516[Abstract/Free Full Text]
15. Epstein RH, Dexter F. Uncertainty in knowing the operating rooms in which cases were performed has little effect on operating room allocations or efficiency. Anesth Analg 2002;95:1726–30[Abstract/Free Full Text]
16. Dexter F, Epstein RH, Traub RD, Xiao Y. Making management decisions on the day of surgery based on operating room efficiency and patient waiting times. Anesthesiology 2004;101:1444–53[Web of Science][Medline]
17. Lee JD, Moray N. Trust, self-confidence, and operators’ adaptation to automation. Int. J Hum Comput Stud 1994;40:153–84
18. Dexter F, Epstein RH, Marcon E, Ledolter J. Estimating the incidence of prolonged turnover times and delays by time of day. Anesthesiology 2005;102:1242–8[Web of Science][Medline]
19. Dexter F, Marcon E, Epstein RH, Ledolter J. Validation of statistical methods to compare cancellation rates on the day of surgery. Anesth Analg 2005;101:465–73[Abstract/Free Full Text]
20. Sandberg WS, Hakkinen M, Egan M, Curran PK, Fairbrother P, Choquette K, Daily B, Sarkka JP, Rattner D. Automatic detection and notification of "wrong patient-wrong location" errors in the operating room. Surg Innov 2005;12:253–60[Abstract/Free Full Text]
21. Egan MT, Sandberg WS. Auto identification technology and its impact on patient safety in the operating room of the future. Surg Innov 2007;14:41–51[Abstract/Free Full Text]
22. Dexter F, Xiao Y, Dow AJ, Strader MM, Ho D, Wachtel RE. Coordination of appointments for anesthesia care outside of operating rooms using an enterprise-wide scheduling system. Anesth Analg 2007;105:1701–10[Abstract/Free Full Text]

This article has been cited by other articles:

				J. J. Pandit and F. Dexter Lack of Sensitivity of Staffing for 8-Hour Sessions to Standard Deviation in Daily Actual Hours of Operating Room Time Used for Surgeons with Long Queues Anesth. Analg., June 1, 2009; 108(6): 1910 - 1915. [Abstract] [Full Text] [PDF]

				F. Dexter, R. H. Epstein, J. D. Lee, and J. Ledolter Automatic Updating of Times Remaining in Surgical Cases Using Bayesian Analysis of Historical Case Duration Data and "Instant Messaging" Updates from Anesthesia Providers Anesth. Analg., March 1, 2009; 108(3): 929 - 940. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Epstein, R. H. Articles by Piotrowski, E. Search for Related Content PubMed PubMed Citation Articles by Epstein, R. H. Articles by Piotrowski, E. Related Collections Economics and Health Care Research