- Email: [email protected]
The effects of an electronic medical record on the completeness of documentation in the anesthesia record
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
journal homepage: www.ijmijournal.com
The effects of an electronic medical record on the completeness of documentation in the anesthesia record Junghwa Jang a , Seung Hum Yu a , Chun-Bae Kim b , Youngkyu Moon c , Sukil Kim d,∗ a
Graduate School of Public Health, Younsei University, 250 Seongsnanno, Seodaemun-Gu, Seoul, Republic of Korea Department of Preventive Medicine, Wonju College of Medicine, Yonsei University, 162 Ilsan-Dong, Wonju-City, Gangwon-Do, Republic of Korea c Graduate School of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul, Republic of Korea d College of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul, Republic of Korea b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objectives: The purpose of this study is to evaluate the completeness of anesthesia recording
Received 14 August 2012
before and after the introduction of an electronic anesthesia record.
Received in revised form
Methods: The study was conducted in a Korean teaching hospital where the EMR was imple-
23 April 2013
mented in October 2008. One hundred paper anesthesia records from July to September 2008
Accepted 24 April 2013
and 150 electronic anesthesia records during the same period in 2009 were randomly sampled. Thirty-four essential items were selected out of all the anesthesia items and grouped
Keywords:
into automatically transferred items and manual entry items. 1, .5 and 0 points were given
Electronic anesthesia record
for each item of complete entry, incomplete entry and no entry respectively. The complete-
Electronic medical record
ness of documentation was defined as the sum of the scores. The influencing factors on the
Completeness of documentation
completeness of documentation were evaluated in total and by the groups. Results: The average completeness score of the electronic anesthesia records was 3.15% higher than that of the paper records. A multiple regression model showed the type of the anesthesia record was a significant factor on the completeness of anesthesia records in all items (ˇ = .98, p < .05) and automatically transferred items (ˇ = .56, p < .01). The type of the anesthesia records had no influence on the completeness in manual entry items. Conclusions: The completeness of an anesthesia record was improved after the implementation of the electronic anesthesia record. The reuse of the data from the EMR was the main contributor to the improved completeness. © 2013 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
The completeness of documentation is important for patient safety [1–3], medico-legal proceedings [4,5] and health insurance reimbursement [6]. Incomplete information on the medical records led the clinicians either to postpone their decisions or to make wrong decisions [2]. The completeness of
anesthesia records is also vital for proper clinical patient documentation [3]. The primary piece of evidence used in court during a malpractice proceeding is the complete documentation [4,5]. Incomplete information on the medical records increases medical liability [7]. The better the completeness of the documentation, the more the clinicians are paid in the situation where the fee-for-service is used as reimbursement schedule like the U.S. and Korea. Even the billing data were
∗ Corresponding author at: Department of Preventive Medicine, College of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul 137-701, Republic of Korea. Tel.: +82 2 2258 7367; fax: +82 2 532 3820. E-mail addresses: [email protected], [email protected] (S. Kim). 1386-5056/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijmedinf.2013.04.004
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
more complete than inpatient data in Canada [8,9] and recommended to use them for statistical purposes. Even though the arguments of the effect of the implementation of EMR continued, the study of the completeness of documentation, the essence of the medical record was not satisfactory. Therefore, it was to be investigated whether the completeness of documentation on an EMR had improved compared with that of the existing paper medical record. This study was carried out on hospital anesthesia records to investigate the completeness of documentation, especially in the operating room in which the medical records should be recorded with speed and accuracy due to urgency. The adoption rate of EMR in Korean teaching hospitals is over 50% [10], and it is interesting how much of an improvement in the completeness of documentation was made by the implementation of EMR compared with the paper medical records.
2.
Methods
2.1.
Subject and data collection method
The study was conducted in a 1200 bed Hospital in Seoul, Korea in which an EMR system was implemented in October 2008. The hospital had been rebuilt just before the implementation of the EMR. This study was conducted by the separate sample pretest–post-test design. One hundred paper anesthesia records were collected out of the patient group who had an operation between July and September of 2008 and 150 electronic anesthesia records during the same period in 2009. The operating rooms were randomly selected matching record period and the first or the last operation in a selected operating room was randomly chosen. The matching ratio was 1.5 based on the increase in the number of general anesthesia of 5352 in 2009 from 3533 in 2008 during the same months of year, while the number of operation room increased to 29 from 17 during the same time periods. The anesthesia records were reviewed from March 10, 2010 to April 10, 2010.
2.2.
Major variables
The difference in the completeness of documentation through the implementation of the EMR was evaluated controlling the recorded period (July, August, September), the recorded time (operation run-time), the record moment (anesthesia start time), and the recorder’s characteristics such as gender and position. The authors built a check list based on Tsai/Bond’s paper [11]. This was modified by comparing the standard anesthesia record provided by the Association of Nurse Anaesthetists of the Korean Nurses Association, the anesthesia records of American Association of Nurse Anaesthetists and the anesthesia records of the study hospital. The check list was finalized through consultation with two professors and three fellows of anesthesiology to fit the study objective and circumstances (Fig. 1). Thirty-four essential items were included in the final version of the check list. Twelve items were automatically transferred from the EMR; registration number, patient’s
703
name, age, gender, surgical department, date of surgery, diagnosis, name of surgical procedure, position, height, weight, and NPO (nothing per oral) status. On the other hand, 22 items were manually recorded by the anesthetist (Fig. 1). For each of the 34 items, 1 point was given if the item was filled in completely, .5 point was given if the item was incomplete, and 0 point if the item was not filled in. The sum of the scores of the items of each record was defined as the completeness score of recording. The records during the period from July to September were selected considering the period for the anesthetists to shift among training wards. Anesthesia start time was assumed to be the time when the anesthesiologist started recording with the classification of four groups as – 08:00–12:00 o’clock, 12:00–16:00 o’clock, 16:00–20:00 o’clock, and 20:00–08:00 o’clock of the next day. Operation run-time was considered as the recording time for an anesthesia record and reckoned as the total duration of anesthesia. It was classified into two categories – under 120 min or above 120 min. An anesthesia record is recorded by the anesthesiologist in charge of the patient under anesthesia for the operation. The characteristic of the recorder was investigated, classifying it into gender and position. The number of the recording anesthesiologists was 57, including specialists and residents.
2.3.
Data analysis
The data was analyzed using SPSS 17.0. The distribution of paper and electronic anesthesia records among 250 sample records was presented with frequency and percentage, and the differences in the distribution of the samples according to the record period, anesthesia start time, operation run-time and the characteristic of the recording anesthesiologists were evaluated through chi-square test. A comparative analysis on the completeness of paper and electronic anesthesia records was done through independent t-test and the difference in the completeness of anesthesia records by the operation runtime was evaluated. The factors influencing the completeness of anesthesia record were evaluated through multiple linear regression analysis.
3.
Results
The record period was evenly distributed among the study months. The greatest proportion of the recording start time based on the anesthesia start time was between 08:00 and 12:00 o’clock in the morning and the number of anesthesia records decreased as time went late. The operation run-time was evenly distributed under and above 120 min (Table 1). The number of the male recorders was higher in electronic anesthesia group to female (p < .01). The composition of the position in each group was different (p < .01). The number of the first year residents occupied 39% of the entire paper anesthesia record group and the fellows occupied 30.7% of the electronic anesthesia group (Table 2).
704
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Fig. 1 – The check list for measuring completeness of documentation in anesthesia records.
The average score of 34 essential items in total was 30.9. The average score of the paper anesthesia records was 30.27 while that of the electronic anesthesia records was 31.34 (p < .01). The average score of 12 automatically transferred items from EMR was higher than that of the paper anesthesia records (p < .01). However, the average scores of 22 items that should have been manually recorded were not statistically different by the type of the anesthesia record (Table 3).
The effect of electronic anesthesia records on the completeness was analyzed through regression models controlling the type of anesthesia record, record period, operation runtime, anesthesia start time, gender and position of the anesthesiologist (Table 4). The result showed that electronic anesthesia records significantly increased completeness of the anesthesia record (p < .05). When the model included automatic transferred items only, the completeness score increased in electronic
705
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Table 1 – Distribution of samples by the type of record unit: number (%). Variables
Type of anesthesia record Paper (n1 = 100)
Record period (month) July August September Anesthesia start time 08–12 12–16 16–20 20–08 Operation run-time <120 min ≥120 min
Total (N = 250)
x2
Electronic (n2 = 150)
35 (35.0) 33 (33.0) 32 (32.0)
53 (35.3) 50 (33.3) 47 (31.3)
88 (35.2) 83 (33.2) 79 (31.6)
.01
54 (54.0) 22 (22.0) 17 (17.0) 7 (7.0)
82 (54.7) 37 (24.7) 24 (16.0) 7 (4.7)
136 (54.4) 59 (23.6) 41 (16.4) 14 (5.6)
.81
47 (47.0) 53 (53.0)
73 (48.7) 77 (51.3)
120 (48.0) 130 (52.0)
.07
Table 2 – Distribution of doctors in charge of recording by the type of record unit: number (%). Variable
Type of anesthesia record Paper (n1 = 100)
Gender Male Female Position Intern 1st year resident 2nd year resident 3rd year resident 4th year resident Fellow Professor a ∗∗∗
Total (N = 250)
x2 /p
Electronic (n2 = 150)
40 (40.0) 60 (60.0)
86 (57.3) 64 (42.7)
126 (50.4) 124 (49.6)
7.21***
2 (2.0) 39 (39.0) 16 (16.0) 11 (11.0) 10 (10.0) 9 (9.0) 13 (13.0)
2 (1.3) 35 (23.3) 17 (11.3) 16 (10.7) 15 (10.0) 46 (30.7) 19 (12.7)
4 (1.6) 74 (29.6) 33 (13.2) 27 (10.8) 25 (10.0) 55 (22.0) 32 (12.8)
.002a
Fisher’s exact test. p < .01.
Table 3 – Completeness of recording of the anesthesia record (mean ± SD). Mode of data acquisitiona
Type of anesthesia record Paper (n1 = 100)
Automatic transfer Manual record Overall a ∗∗∗
11.08 ± .97 19.41 ± 1.82 30.27 ± 3.02
T value
11.42 ± .87 19.65 ± 2.06 30.91 ± 3.02
5.02*** 1.61 2.78***
Electronic (n2 = 150) 11.65 ± .71 19.82 ± 2.19 31.34 ± 2.96
Refer to Fig. 1. p < .01.
anesthesia record (p < .01). However, the completeness score was decreased in September and when written from 20:00 to 08:00 o’clock (<.05). For the manually acquired items, the type of anesthesia record had no effect on the completeness score. However, the operation run-time strongly increased and anesthesia start time between 20:00 and 08:00 o’clock strongly decreased the completeness (p < .01).
4.
Total (N = 250)
Discussion
During the building process of new electronic medical record, some of the entry items were added to the legacy paper anesthesia record. For example, there were only 8 items for the blood test results and the rest of the items were
added by writing the test name and value by hand in the paper anesthesia record. The doctors usually recorded the abnormal values only and not the normal values. In the electronic anesthesia records it was easy to identify the missing entries and even the normal values were entered again to complete items required. Therefore only 34 essential items which must be recorded in an anesthesia record were selected for a valid comparison between the paper and electronic anesthesia records and other items were excluded from the check list. According to an anesthesia textbook [12], an anesthesia record should contain every event which occurred during a surgical and anesthesia procedure by the anesthesiologist, and a patient’s condition to such a degree that other anesthesiologists could carry out the anesthesia procedure in the same way. This study would become more valid if more
706
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Table 4 – Effect of electronic anesthesia record on the completeness of anesthesia record. Mode of data acquisition (ˇ)a
Factors Automatic transfer Type of anesthesia record (paper = 0) Electronic Record period (July = 0) August September Operation run-time Anesthesia start time (08–12 o’clock = 0) 12–16 o’clock 16–20 o’clock 20–08 o’clock Gender of the anesthesiologist (male = 0) Female Position of the anesthesiologist R2 a ∗ ∗∗∗
Manual record
.56***
.43
Overall .98*
−.10 −.31* .15
−.03 −.43 1.26***
.22 −.14 .14
.001 .05 −.50*
−.38 .21 −2.43***
−.25 .44 −.22
−.03 −.01 .15
.04 −.05 .19
−.67 −.02 .05
Refer to Fig. 1. p < .05. p < .01.
items in addition to 34 essential ones were included in the analysis. The study result showed that the completeness of the electronic anesthesia records was 1.07 points (3.15%) higher than that of the paper anesthesia records. We considered 3 reasons for the high completeness. First, it was easy to identify the missing items in electronic anesthesia record and gave more possibility for users to fill them out. Second, the check boxes increased the completeness of documentation even in the normal value items [13]. Finally, items for patient’s basic information were automatically transferred from EMR [2] minimizing the missing items. Even though the type of anesthesia record was a significant factor influencing the completeness of anesthesia records in the regression model, the R2 was only 5%. The model was re-analyzed by way of data gathering, automatic transfer items and manual entry items. The R2 of automatic transfer increased to 15.4% and that of manual entry increased to 18.9%. The type of data gathering was the major factor that can increase the explanation power of the model. The operations started from 08:00 p.m. to 08:00 a.m. next day were generally emergencies. The night duty anesthesiologists were normally lower year residents with little experience and the recordings had more missing items due to their tiredness and reduced efficiency [14,15]. This could explain the reason why completeness of records from 08:00 p.m. to 08:00 a.m. next day was significantly lower than that from 08:00 a.m. to 12:00 a.m. The rotation of anesthesiologists among training hospitals every September in the study hospital was reflected in the result that the completeness score in September was significantly lower than that of July. The completeness score of the operation run-time more than 2 h was 1.23 point higher than that of operation runtime less than 2 h in the manual entry items. Even though the basic anesthesia procedures and the number of basic items to be recorded were the same in all the operations, the short operation time is considered to increase the possibility of incomplete recording.
We reviewed research subjects and methods in several studies on the completeness of documents in EMR. There was a study in which the completeness of documentation was focused on the billing purpose [7]. Other studies were based on questionnaire answered by clinicians [16], qualitative analysis [17] or subject quantification by reviewers [2]. Our study is unique comparing to those studies because we introduced objectively quantitative approach to evaluate the completeness of documentation and found same result of increase in the completeness of documentation in general [2,7,16,17] and due to automatic transfer of previously entered data [2]. The limitations in general could be considered, given the setting of the study as a single university hospital which implemented an EMR recently in a new building. The EMR was developed in house as other large scale university hospitals in Korea. The level of sophistication in EMR and electronic anesthesia record may be different from other hospitals with similar scale or from smaller scale hospitals which are using vendor developed systems. More participating hospitals are required to increase the generalization and comparativeness of the study results. With the recent development of anesthesia devices and information and communication technology, data capture from the devices to electronic anesthesia record is getting more common. And there are such efforts as to increase completeness by developing better user interface [18]. However, they were not under consideration because the electronic anesthesia record of the study hospital was not equipped with this kind of functions.
5.
Conclusion
The completeness of anesthesia documentation was improved after the introduction of electronic anesthesia record. The reuse of same items from EMR was the major reason of the improvement. The completeness of the record in this study was focused on the presence of 34 items of the anesthesia records. Further
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Summary points What was already known on this topic • The completeness of documents in EMR was increased when it was measure for were billing purposes or by questionnaire answered by clinicians, qualitative analysis or subject quantification by reviewers. • The reason of increasing completeness of documentation was automatic transfer of previously entered data. What this study has added to our knowledge • The completeness of documentation of electronic anesthesia record was improved when measured by objectively quantitative approach. • There was no improvement of the completeness of documentation in manually recorded items of electronic anesthesia record. • The reuse of same items from EMR was the major reason why the completeness of electronic anesthesia record was improved.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
studies are needed to evaluate other aspects of the completeness of documentation in EMR such as the quantity of the recorded information, the validity of the recorded information, and the consistency with other records.
[9]
[10]
Authors’ contribution Junghwa Jang is the first author of the paper. She reviewed related papers, built the research hypothesis, gathered data wrote most part of the discussion. Seung Hum Yu and Chun-Bae Kim gave their important comment on the research design and wrote some part of the discussion. Dr. Kim also reviews the English translation. Youngkyu Moon analyzed the data and contributed to the interpretation of the result. He participated in the translation of the paper into English. Sukil Kim is the corresponding author. He also reviewed related papers, help Ms. Jang to build up the research hypothesis, review the analysis result, wrote some part of the discussion and edited English version of the paper.
Competing interest
[11]
[12] [13]
[14]
[15]
[16]
None declared.
[17]
references
[18]
[1] L.L. Leape, D.M. Berwick, D.W. Bates, What practices will most improve safety? Evidence-based medicine meets
707
patient safety, J. Am. Med. Assoc. 288 (July (4)) (2002) 501–507 (Editorial Research Support, Non-U. S. Gov’t). P.C. Tang, M.P. LaRosa, S.M. Gorden, Use of computer-based records, completeness of documentation, and appropriateness of documented clinical decisions, J. Am. Med. Inform. Assoc. 6 (May–June (3)) (1999) 245–251. M.B. Flynn, D.A. Allen, The operative note as billing documentation: a preliminary report, Am. Surg. (Eval. Stud.) 70 (July (7)) (2004) 570–574 (discussion 4-5). J.M. Feldman, Do anesthesia information systems increase malpractice exposure? Results of a survey, Anesth. Analg. 99 (September (3)) (2004) 840–843 (table of contents). P.S. Schoenfeld, M.D. Baker, Documentation in the pediatric emergency department: a review of resuscitation cases, Ann. Emerg. Med. 20 (June (6)) (1991) 641–643. S. Kheterpal, R. Gupta, J.M. Blum, K.K. Tremper, M. O’Reilly, P.E. Kazanjian, Electronic reminders improve procedure documentation compliance and professional fee reimbursement, Anesth. Analg. 104 (March (3)) (2007) 592–597 (Randomized Controlled Trial). M.M. Vigoda, D.A. Lubarsky, 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. 103 (July (1)) (2006) 131–136. X. Li, C. King, C. deGara, J. White, M. Winget, Validation of colorectal cancer surgery data from administrative data sources, BMC Med. Res. Methodol. 12 (2012) 97. X. Li, R. Hilsden, S. Hossain, J. Fleming, M. Winget, Validation of administrative data sources for endoscopy utilization in colorectal cancer diagnosis, BMC Health Serv. Res. 12 (2012) 358. D. Yoon, B.C. Chang, S.W. Kang, H. Bae, R.W. Park, Adoption of electronic health records in Korean tertiary teaching and general hospitals, Int. J. Med. Inf. 81 (March (3)) (2012) 196–203. J. Tsai, G. Bond, A comparison of electronic records to paper records in mental health centers, Int. J. Qual. Health Care 20 (April (2)) (2008) 136–143. G. Morgan Jr., M. Mikhail, M. Murray, Clinical Anesthesiology, McGraw-Hill, Columbus, Ohio, 2006. J. Roukema, R.K. Los, S.E. Bleeker, A.M. van Ginneken, J. van der Lei, H.A. Moll, Paper versus computer: feasibility of an electronic medical record in general pediatrics, Pediatrics 117 (January (1)) (2006) 15–21. P. Gander, H. Purnell, A. Garden, A. Woodward, Work patterns and fatigue-related risk among junior doctors, Occup. Environ. Med. 64 (November (11)) (2007) 733–738. R. Smith-Coggins, M.R. Rosekind, S. Hurd, K.R. Buccino, Relationship of day versus night sleep to physician performance and mood, Ann. Emerg. Med. 24 (November (5)) (1994) 928–934. D.B. Hier, A. Rothschild, A. LeMaistre, J. Keeler, Differing faculty and housestaff acceptance of an electronic health record, Int. J. Med. Inf. 74 (August (7/8)) (2005) 657–662. R.H. Miller, I. Sim, Physicians’ use of electronic medical records: barriers and solutions, Health Aff. (Millwood) 23 (March–April (2)) (2004) 116–126. A. Avidan, C. Weissman, Record completeness and data concordance in an anesthesia information management system using context-sensitive mandatory data-entry fields, Int. J. Med. Inf. 81 (March (3)) (2012) 173–181.
journal homepage: www.ijmijournal.com
The effects of an electronic medical record on the completeness of documentation in the anesthesia record Junghwa Jang a , Seung Hum Yu a , Chun-Bae Kim b , Youngkyu Moon c , Sukil Kim d,∗ a
Graduate School of Public Health, Younsei University, 250 Seongsnanno, Seodaemun-Gu, Seoul, Republic of Korea Department of Preventive Medicine, Wonju College of Medicine, Yonsei University, 162 Ilsan-Dong, Wonju-City, Gangwon-Do, Republic of Korea c Graduate School of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul, Republic of Korea d College of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul, Republic of Korea b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objectives: The purpose of this study is to evaluate the completeness of anesthesia recording
Received 14 August 2012
before and after the introduction of an electronic anesthesia record.
Received in revised form
Methods: The study was conducted in a Korean teaching hospital where the EMR was imple-
23 April 2013
mented in October 2008. One hundred paper anesthesia records from July to September 2008
Accepted 24 April 2013
and 150 electronic anesthesia records during the same period in 2009 were randomly sampled. Thirty-four essential items were selected out of all the anesthesia items and grouped
Keywords:
into automatically transferred items and manual entry items. 1, .5 and 0 points were given
Electronic anesthesia record
for each item of complete entry, incomplete entry and no entry respectively. The complete-
Electronic medical record
ness of documentation was defined as the sum of the scores. The influencing factors on the
Completeness of documentation
completeness of documentation were evaluated in total and by the groups. Results: The average completeness score of the electronic anesthesia records was 3.15% higher than that of the paper records. A multiple regression model showed the type of the anesthesia record was a significant factor on the completeness of anesthesia records in all items (ˇ = .98, p < .05) and automatically transferred items (ˇ = .56, p < .01). The type of the anesthesia records had no influence on the completeness in manual entry items. Conclusions: The completeness of an anesthesia record was improved after the implementation of the electronic anesthesia record. The reuse of the data from the EMR was the main contributor to the improved completeness. © 2013 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
The completeness of documentation is important for patient safety [1–3], medico-legal proceedings [4,5] and health insurance reimbursement [6]. Incomplete information on the medical records led the clinicians either to postpone their decisions or to make wrong decisions [2]. The completeness of
anesthesia records is also vital for proper clinical patient documentation [3]. The primary piece of evidence used in court during a malpractice proceeding is the complete documentation [4,5]. Incomplete information on the medical records increases medical liability [7]. The better the completeness of the documentation, the more the clinicians are paid in the situation where the fee-for-service is used as reimbursement schedule like the U.S. and Korea. Even the billing data were
∗ Corresponding author at: Department of Preventive Medicine, College of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho-Gu, Seoul 137-701, Republic of Korea. Tel.: +82 2 2258 7367; fax: +82 2 532 3820. E-mail addresses: [email protected], [email protected] (S. Kim). 1386-5056/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijmedinf.2013.04.004
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
more complete than inpatient data in Canada [8,9] and recommended to use them for statistical purposes. Even though the arguments of the effect of the implementation of EMR continued, the study of the completeness of documentation, the essence of the medical record was not satisfactory. Therefore, it was to be investigated whether the completeness of documentation on an EMR had improved compared with that of the existing paper medical record. This study was carried out on hospital anesthesia records to investigate the completeness of documentation, especially in the operating room in which the medical records should be recorded with speed and accuracy due to urgency. The adoption rate of EMR in Korean teaching hospitals is over 50% [10], and it is interesting how much of an improvement in the completeness of documentation was made by the implementation of EMR compared with the paper medical records.
2.
Methods
2.1.
Subject and data collection method
The study was conducted in a 1200 bed Hospital in Seoul, Korea in which an EMR system was implemented in October 2008. The hospital had been rebuilt just before the implementation of the EMR. This study was conducted by the separate sample pretest–post-test design. One hundred paper anesthesia records were collected out of the patient group who had an operation between July and September of 2008 and 150 electronic anesthesia records during the same period in 2009. The operating rooms were randomly selected matching record period and the first or the last operation in a selected operating room was randomly chosen. The matching ratio was 1.5 based on the increase in the number of general anesthesia of 5352 in 2009 from 3533 in 2008 during the same months of year, while the number of operation room increased to 29 from 17 during the same time periods. The anesthesia records were reviewed from March 10, 2010 to April 10, 2010.
2.2.
Major variables
The difference in the completeness of documentation through the implementation of the EMR was evaluated controlling the recorded period (July, August, September), the recorded time (operation run-time), the record moment (anesthesia start time), and the recorder’s characteristics such as gender and position. The authors built a check list based on Tsai/Bond’s paper [11]. This was modified by comparing the standard anesthesia record provided by the Association of Nurse Anaesthetists of the Korean Nurses Association, the anesthesia records of American Association of Nurse Anaesthetists and the anesthesia records of the study hospital. The check list was finalized through consultation with two professors and three fellows of anesthesiology to fit the study objective and circumstances (Fig. 1). Thirty-four essential items were included in the final version of the check list. Twelve items were automatically transferred from the EMR; registration number, patient’s
703
name, age, gender, surgical department, date of surgery, diagnosis, name of surgical procedure, position, height, weight, and NPO (nothing per oral) status. On the other hand, 22 items were manually recorded by the anesthetist (Fig. 1). For each of the 34 items, 1 point was given if the item was filled in completely, .5 point was given if the item was incomplete, and 0 point if the item was not filled in. The sum of the scores of the items of each record was defined as the completeness score of recording. The records during the period from July to September were selected considering the period for the anesthetists to shift among training wards. Anesthesia start time was assumed to be the time when the anesthesiologist started recording with the classification of four groups as – 08:00–12:00 o’clock, 12:00–16:00 o’clock, 16:00–20:00 o’clock, and 20:00–08:00 o’clock of the next day. Operation run-time was considered as the recording time for an anesthesia record and reckoned as the total duration of anesthesia. It was classified into two categories – under 120 min or above 120 min. An anesthesia record is recorded by the anesthesiologist in charge of the patient under anesthesia for the operation. The characteristic of the recorder was investigated, classifying it into gender and position. The number of the recording anesthesiologists was 57, including specialists and residents.
2.3.
Data analysis
The data was analyzed using SPSS 17.0. The distribution of paper and electronic anesthesia records among 250 sample records was presented with frequency and percentage, and the differences in the distribution of the samples according to the record period, anesthesia start time, operation run-time and the characteristic of the recording anesthesiologists were evaluated through chi-square test. A comparative analysis on the completeness of paper and electronic anesthesia records was done through independent t-test and the difference in the completeness of anesthesia records by the operation runtime was evaluated. The factors influencing the completeness of anesthesia record were evaluated through multiple linear regression analysis.
3.
Results
The record period was evenly distributed among the study months. The greatest proportion of the recording start time based on the anesthesia start time was between 08:00 and 12:00 o’clock in the morning and the number of anesthesia records decreased as time went late. The operation run-time was evenly distributed under and above 120 min (Table 1). The number of the male recorders was higher in electronic anesthesia group to female (p < .01). The composition of the position in each group was different (p < .01). The number of the first year residents occupied 39% of the entire paper anesthesia record group and the fellows occupied 30.7% of the electronic anesthesia group (Table 2).
704
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Fig. 1 – The check list for measuring completeness of documentation in anesthesia records.
The average score of 34 essential items in total was 30.9. The average score of the paper anesthesia records was 30.27 while that of the electronic anesthesia records was 31.34 (p < .01). The average score of 12 automatically transferred items from EMR was higher than that of the paper anesthesia records (p < .01). However, the average scores of 22 items that should have been manually recorded were not statistically different by the type of the anesthesia record (Table 3).
The effect of electronic anesthesia records on the completeness was analyzed through regression models controlling the type of anesthesia record, record period, operation runtime, anesthesia start time, gender and position of the anesthesiologist (Table 4). The result showed that electronic anesthesia records significantly increased completeness of the anesthesia record (p < .05). When the model included automatic transferred items only, the completeness score increased in electronic
705
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Table 1 – Distribution of samples by the type of record unit: number (%). Variables
Type of anesthesia record Paper (n1 = 100)
Record period (month) July August September Anesthesia start time 08–12 12–16 16–20 20–08 Operation run-time <120 min ≥120 min
Total (N = 250)
x2
Electronic (n2 = 150)
35 (35.0) 33 (33.0) 32 (32.0)
53 (35.3) 50 (33.3) 47 (31.3)
88 (35.2) 83 (33.2) 79 (31.6)
.01
54 (54.0) 22 (22.0) 17 (17.0) 7 (7.0)
82 (54.7) 37 (24.7) 24 (16.0) 7 (4.7)
136 (54.4) 59 (23.6) 41 (16.4) 14 (5.6)
.81
47 (47.0) 53 (53.0)
73 (48.7) 77 (51.3)
120 (48.0) 130 (52.0)
.07
Table 2 – Distribution of doctors in charge of recording by the type of record unit: number (%). Variable
Type of anesthesia record Paper (n1 = 100)
Gender Male Female Position Intern 1st year resident 2nd year resident 3rd year resident 4th year resident Fellow Professor a ∗∗∗
Total (N = 250)
x2 /p
Electronic (n2 = 150)
40 (40.0) 60 (60.0)
86 (57.3) 64 (42.7)
126 (50.4) 124 (49.6)
7.21***
2 (2.0) 39 (39.0) 16 (16.0) 11 (11.0) 10 (10.0) 9 (9.0) 13 (13.0)
2 (1.3) 35 (23.3) 17 (11.3) 16 (10.7) 15 (10.0) 46 (30.7) 19 (12.7)
4 (1.6) 74 (29.6) 33 (13.2) 27 (10.8) 25 (10.0) 55 (22.0) 32 (12.8)
.002a
Fisher’s exact test. p < .01.
Table 3 – Completeness of recording of the anesthesia record (mean ± SD). Mode of data acquisitiona
Type of anesthesia record Paper (n1 = 100)
Automatic transfer Manual record Overall a ∗∗∗
11.08 ± .97 19.41 ± 1.82 30.27 ± 3.02
T value
11.42 ± .87 19.65 ± 2.06 30.91 ± 3.02
5.02*** 1.61 2.78***
Electronic (n2 = 150) 11.65 ± .71 19.82 ± 2.19 31.34 ± 2.96
Refer to Fig. 1. p < .01.
anesthesia record (p < .01). However, the completeness score was decreased in September and when written from 20:00 to 08:00 o’clock (<.05). For the manually acquired items, the type of anesthesia record had no effect on the completeness score. However, the operation run-time strongly increased and anesthesia start time between 20:00 and 08:00 o’clock strongly decreased the completeness (p < .01).
4.
Total (N = 250)
Discussion
During the building process of new electronic medical record, some of the entry items were added to the legacy paper anesthesia record. For example, there were only 8 items for the blood test results and the rest of the items were
added by writing the test name and value by hand in the paper anesthesia record. The doctors usually recorded the abnormal values only and not the normal values. In the electronic anesthesia records it was easy to identify the missing entries and even the normal values were entered again to complete items required. Therefore only 34 essential items which must be recorded in an anesthesia record were selected for a valid comparison between the paper and electronic anesthesia records and other items were excluded from the check list. According to an anesthesia textbook [12], an anesthesia record should contain every event which occurred during a surgical and anesthesia procedure by the anesthesiologist, and a patient’s condition to such a degree that other anesthesiologists could carry out the anesthesia procedure in the same way. This study would become more valid if more
706
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Table 4 – Effect of electronic anesthesia record on the completeness of anesthesia record. Mode of data acquisition (ˇ)a
Factors Automatic transfer Type of anesthesia record (paper = 0) Electronic Record period (July = 0) August September Operation run-time Anesthesia start time (08–12 o’clock = 0) 12–16 o’clock 16–20 o’clock 20–08 o’clock Gender of the anesthesiologist (male = 0) Female Position of the anesthesiologist R2 a ∗ ∗∗∗
Manual record
.56***
.43
Overall .98*
−.10 −.31* .15
−.03 −.43 1.26***
.22 −.14 .14
.001 .05 −.50*
−.38 .21 −2.43***
−.25 .44 −.22
−.03 −.01 .15
.04 −.05 .19
−.67 −.02 .05
Refer to Fig. 1. p < .05. p < .01.
items in addition to 34 essential ones were included in the analysis. The study result showed that the completeness of the electronic anesthesia records was 1.07 points (3.15%) higher than that of the paper anesthesia records. We considered 3 reasons for the high completeness. First, it was easy to identify the missing items in electronic anesthesia record and gave more possibility for users to fill them out. Second, the check boxes increased the completeness of documentation even in the normal value items [13]. Finally, items for patient’s basic information were automatically transferred from EMR [2] minimizing the missing items. Even though the type of anesthesia record was a significant factor influencing the completeness of anesthesia records in the regression model, the R2 was only 5%. The model was re-analyzed by way of data gathering, automatic transfer items and manual entry items. The R2 of automatic transfer increased to 15.4% and that of manual entry increased to 18.9%. The type of data gathering was the major factor that can increase the explanation power of the model. The operations started from 08:00 p.m. to 08:00 a.m. next day were generally emergencies. The night duty anesthesiologists were normally lower year residents with little experience and the recordings had more missing items due to their tiredness and reduced efficiency [14,15]. This could explain the reason why completeness of records from 08:00 p.m. to 08:00 a.m. next day was significantly lower than that from 08:00 a.m. to 12:00 a.m. The rotation of anesthesiologists among training hospitals every September in the study hospital was reflected in the result that the completeness score in September was significantly lower than that of July. The completeness score of the operation run-time more than 2 h was 1.23 point higher than that of operation runtime less than 2 h in the manual entry items. Even though the basic anesthesia procedures and the number of basic items to be recorded were the same in all the operations, the short operation time is considered to increase the possibility of incomplete recording.
We reviewed research subjects and methods in several studies on the completeness of documents in EMR. There was a study in which the completeness of documentation was focused on the billing purpose [7]. Other studies were based on questionnaire answered by clinicians [16], qualitative analysis [17] or subject quantification by reviewers [2]. Our study is unique comparing to those studies because we introduced objectively quantitative approach to evaluate the completeness of documentation and found same result of increase in the completeness of documentation in general [2,7,16,17] and due to automatic transfer of previously entered data [2]. The limitations in general could be considered, given the setting of the study as a single university hospital which implemented an EMR recently in a new building. The EMR was developed in house as other large scale university hospitals in Korea. The level of sophistication in EMR and electronic anesthesia record may be different from other hospitals with similar scale or from smaller scale hospitals which are using vendor developed systems. More participating hospitals are required to increase the generalization and comparativeness of the study results. With the recent development of anesthesia devices and information and communication technology, data capture from the devices to electronic anesthesia record is getting more common. And there are such efforts as to increase completeness by developing better user interface [18]. However, they were not under consideration because the electronic anesthesia record of the study hospital was not equipped with this kind of functions.
5.
Conclusion
The completeness of anesthesia documentation was improved after the introduction of electronic anesthesia record. The reuse of same items from EMR was the major reason of the improvement. The completeness of the record in this study was focused on the presence of 34 items of the anesthesia records. Further
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 2 ( 2 0 1 3 ) 702–707
Summary points What was already known on this topic • The completeness of documents in EMR was increased when it was measure for were billing purposes or by questionnaire answered by clinicians, qualitative analysis or subject quantification by reviewers. • The reason of increasing completeness of documentation was automatic transfer of previously entered data. What this study has added to our knowledge • The completeness of documentation of electronic anesthesia record was improved when measured by objectively quantitative approach. • There was no improvement of the completeness of documentation in manually recorded items of electronic anesthesia record. • The reuse of same items from EMR was the major reason why the completeness of electronic anesthesia record was improved.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
studies are needed to evaluate other aspects of the completeness of documentation in EMR such as the quantity of the recorded information, the validity of the recorded information, and the consistency with other records.
[9]
[10]
Authors’ contribution Junghwa Jang is the first author of the paper. She reviewed related papers, built the research hypothesis, gathered data wrote most part of the discussion. Seung Hum Yu and Chun-Bae Kim gave their important comment on the research design and wrote some part of the discussion. Dr. Kim also reviews the English translation. Youngkyu Moon analyzed the data and contributed to the interpretation of the result. He participated in the translation of the paper into English. Sukil Kim is the corresponding author. He also reviewed related papers, help Ms. Jang to build up the research hypothesis, review the analysis result, wrote some part of the discussion and edited English version of the paper.
Competing interest
[11]
[12] [13]
[14]
[15]
[16]
None declared.
[17]
references
[18]
[1] L.L. Leape, D.M. Berwick, D.W. Bates, What practices will most improve safety? Evidence-based medicine meets
707
patient safety, J. Am. Med. Assoc. 288 (July (4)) (2002) 501–507 (Editorial Research Support, Non-U. S. Gov’t). P.C. Tang, M.P. LaRosa, S.M. Gorden, Use of computer-based records, completeness of documentation, and appropriateness of documented clinical decisions, J. Am. Med. Inform. Assoc. 6 (May–June (3)) (1999) 245–251. M.B. Flynn, D.A. Allen, The operative note as billing documentation: a preliminary report, Am. Surg. (Eval. Stud.) 70 (July (7)) (2004) 570–574 (discussion 4-5). J.M. Feldman, Do anesthesia information systems increase malpractice exposure? Results of a survey, Anesth. Analg. 99 (September (3)) (2004) 840–843 (table of contents). P.S. Schoenfeld, M.D. Baker, Documentation in the pediatric emergency department: a review of resuscitation cases, Ann. Emerg. Med. 20 (June (6)) (1991) 641–643. S. Kheterpal, R. Gupta, J.M. Blum, K.K. Tremper, M. O’Reilly, P.E. Kazanjian, Electronic reminders improve procedure documentation compliance and professional fee reimbursement, Anesth. Analg. 104 (March (3)) (2007) 592–597 (Randomized Controlled Trial). M.M. Vigoda, D.A. Lubarsky, 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. 103 (July (1)) (2006) 131–136. X. Li, C. King, C. deGara, J. White, M. Winget, Validation of colorectal cancer surgery data from administrative data sources, BMC Med. Res. Methodol. 12 (2012) 97. X. Li, R. Hilsden, S. Hossain, J. Fleming, M. Winget, Validation of administrative data sources for endoscopy utilization in colorectal cancer diagnosis, BMC Health Serv. Res. 12 (2012) 358. D. Yoon, B.C. Chang, S.W. Kang, H. Bae, R.W. Park, Adoption of electronic health records in Korean tertiary teaching and general hospitals, Int. J. Med. Inf. 81 (March (3)) (2012) 196–203. J. Tsai, G. Bond, A comparison of electronic records to paper records in mental health centers, Int. J. Qual. Health Care 20 (April (2)) (2008) 136–143. G. Morgan Jr., M. Mikhail, M. Murray, Clinical Anesthesiology, McGraw-Hill, Columbus, Ohio, 2006. J. Roukema, R.K. Los, S.E. Bleeker, A.M. van Ginneken, J. van der Lei, H.A. Moll, Paper versus computer: feasibility of an electronic medical record in general pediatrics, Pediatrics 117 (January (1)) (2006) 15–21. P. Gander, H. Purnell, A. Garden, A. Woodward, Work patterns and fatigue-related risk among junior doctors, Occup. Environ. Med. 64 (November (11)) (2007) 733–738. R. Smith-Coggins, M.R. Rosekind, S. Hurd, K.R. Buccino, Relationship of day versus night sleep to physician performance and mood, Ann. Emerg. Med. 24 (November (5)) (1994) 928–934. D.B. Hier, A. Rothschild, A. LeMaistre, J. Keeler, Differing faculty and housestaff acceptance of an electronic health record, Int. J. Med. Inf. 74 (August (7/8)) (2005) 657–662. R.H. Miller, I. Sim, Physicians’ use of electronic medical records: barriers and solutions, Health Aff. (Millwood) 23 (March–April (2)) (2004) 116–126. A. Avidan, C. Weissman, Record completeness and data concordance in an anesthesia information management system using context-sensitive mandatory data-entry fields, Int. J. Med. Inf. 81 (March (3)) (2012) 173–181.