- Email: [email protected]
An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience
G Model
ACCPM-104; No. of Pages 5 Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
Original Article
An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience Cyrus Motamed *, Jean Louis Bourgain Department of Anaesthesia, Gustave-Roussy Cancer Institute, 114, rue E´douard-Vaillant, 94805 Villejuif cedex, France
A R T I C L E I N F O
A B S T R A C T
Article history: Available online xxx
Introduction: Anaesthesia Information Management Systems (AIMS) generate large amounts of data, which might be useful for quality assurance programs. This study was designed to highlight the multiple contributions of our AIMS system in extracting quality indicators over a period of 10 years. Methods: The study was conducted from 2002 to 2011. Two methods were used to extract anaesthesia indicators: the manual extraction of individual files for monitoring neuromuscular relaxation and structured query language (SQL) extraction for other indicators which were postoperative nausea and vomiting (PONV), pain, sedation scores, pain-related medications, scores and postoperative hypothermia. For each indicator, a program of information/meetings and adaptation/suggestions for operating room and PACU personnel was initiated to improve quality assurance, while data were extracted each year. Results: The study included 77,573 patients. The mean overall completeness of data for the initial years ranged from 55 to 85% and was indicator-dependent, which then improved to 95% completeness for the last 5 years. The incidence of neuromuscular monitoring was initially 67% and then increased to 95% (P < 0.05). The rate of pharmacological reversal remained around 53% throughout the study. Regarding SQL data, an improvement of severe postoperative pain and PONV scores was observed throughout the study, while mild postoperative hypothermia remained a challenge, despite efforts for improvement. Discussion: The AIMS system permitted the follow-up of certain indicators through manual sampling and many more via SQL extraction in a sustained and non-time-consuming way across years. However, it requires competent and especially dedicated resources to handle the database. ß 2016 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Keywords: AIMS Quality assurance in anaesthesia Indicators of quality Assurance
1. Introduction The computerized recording of anaesthesia charts, or Anaesthesia Information Management Systems (AIMS), has been available for anaesthesia providers for over two decades [1]. They were originally designed to replace traditional, voluntary, hand recording systems [2]. Automated anaesthesia records and computerized data are described as being more accurate since the phenomena of smoothing, or resistance to the reporting of extreme values, cannot be present [3,4] and, at the least, they create more time for patient-surveillance. Both systems lack total accuracy. For example, the AIMS system does not rely solely on automated parameter recording since self-reporting (e.g.: adverse events) remains necessary and might be under-reported or less accurate [3,5]. In parallel, this information technology generates * Corresponding author. E-mail address: [email protected] (C. Motamed).
huge databases, which allow anaesthesia providers to rapidly consult each anaesthesia record, thus avoiding the waiting time associated with paper record retrieval and transfer. However, data extraction by SQL (Structured Query Language) is also available and can be used to monitor certain data categories, such as intra- or postoperative immediate outcomes [6]. The effective monitoring of medical activity is an integral part of any assurance quality program, together with sustained control of multiple indicators. We previously published a brief report on the overall trends in the main indicators of our quality assurance program for 9 consecutive years [7]. The primary objective of the present study was to quantify the effect of several interventions or changes in practice during 10 consecutive years in our anaesthesia department. 2. Patients and methods In March 2001, our operating rooms were equipped with the ARKTM system integrated with ADU S/5 workstations
http://dx.doi.org/10.1016/j.accpm.2015.11.002 2352-5568/ß 2016 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 2
C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
(Datex-Ohmeda General Electric) for the computerized recording of anaesthesia records and recognized as an AIMS. In order to allow for a washout period during which staff (including users and administrators) became collectively familiar with the system, the actual study period ran from 2002 to 2011. While interventions by anaesthesia providers are indexed manually, physiologic data and vital signs are recorded automatically (e.g.: EKG, pulse, blood pressure, temperature, ventilatory parameters related to anaesthesia machines and inhalational gas consumption). All patients who had anaesthesia, be it general, sedation, regional or local with anaesthesia surveillance, were considered. Upon arrival in the Postoperative Anaesthesia Care Unit (PACU), several clinical items must be manually recorded into the system (e.g.: respiratory status, pain and postoperative nausea and vomiting [PONV], sedation scores and their respective related medications, or central temperature). Our institution’s local ethics committee authorized our department to extract information from the database for our quality assurance program and subsequently publish general results (Avis no 92012/33465). Our clinical procedures strongly encourage anaesthesia providers to index all non-automated administrative, clinical and pharmacological interventions in order to decrease the percentage of incomplete files. During this period, our quality management team started concomitantly monitoring anaesthesia care indicators via manual searches using the PDF storage database (personal anaesthesia file for each patient, Archive Browser1) and by SQL extractions over the entire database for specific indicators. All items extracted by SQL analysis were manually verified for a small number of patients during a short period of time in order to rule out duplicate or missing files. For the purposes of quality assurance in our anaesthesia department, we selected several indicators for extraction: the percentage of pharmacological reversals and instrumental monitoring of neuromuscular blockades, opioid consumption for the intraoperative period and while in the post-anaesthetic care unit (PACU). Other extracted indicators included: pain scores upon arrival, morphine and other analgesic drug consumption, sedation scores, central temperature (tympanic) and PONV scores, which are all manually indexed by the PACU personnel. Results were analysed and meetings on morbidity/mortality were organized in order to issue suggestions for improvements. Indicators were therefore re-checked in further assessments. The suggested interventions and changes in practice were: muscle relaxants: the mandatory use of our neuromuscular monitoring system NMT Datex-Ohmeda1, which is connected to our anaesthesia station and the AIMS system and reversal of neuromuscular blockade and extubation according to updated general clinical guidelines; intra- and postoperative pain management was initially guided by promoting multimodal non-opioid analgesia, and the decrease of postoperative morphine, epidural analgesia using local anaesthetics for major thoracoabdominal surgery, and the progressive shift to use of remifentanil for early extubation; postoperative nausea and vomiting was initially managed at the discretion of the anaesthesiologists. However, after a primary evaluation in 2005 [8], a new protocol using intraoperative dexamethasone and droperidol with odansetron in the PACU was initiated in 2006. This protocol was then readjusted in 2012; postoperative inadvertent hypothermia was discussed every 2 years not only with anaesthesia providers, but also with other operating room actors. Guidelines were issued concerning the management of room temperature, monitoring of intraoperative temperature, and blankets to be available for the patients before entering the operating room area, and the availability of an adequate number of active warming devices.
All data for one patient are recorded as a PDF file, which is printed and added to his/her hospital file when leaving the PACU. The database is also searchable by keywords for system and quality assurance administrators; however, data extraction is not intuitive and requires several hours of training, which can be considered as a limitation of the system. Items, such as the presence and quality of neuromuscular blockade monitoring data (when SQL extraction was not possible), were extracted manually by sampling through a fixed period and for a limited number of patients (200). 2.1. Statistical analysis Results are presented as percentages or means standard deviations when appropriate. Because of the data distribution and the number of missing files in the initial period only, descriptive numbers with their trends are presented. Comparative statistics were performed for a limited number of patients as concerns muscle relaxant monitoring, or for years in which missing files were less than 10% by using ANOVA or Chi2 tests as appropriate. A P value of 0.05 was considered to be statistically significant. 3. Results The data were collected over 10 consecutive years from 2002 to 2011. The total number of patients registered for this period was 77,573. The mean completeness or exhaustivity of the data ranged from 90% in 2002 to > 95% in 2011. However, this was item-dependent; the worst case was the PONV score, which started as low as 50% and attained 86% in 2011. 3.1. Monitoring neuromuscular blockades Although the percentage of reversal could be rapidly accessed by SQL (51 to 57% across years), the incidence of instrumental monitoring could not be extracted from the database and required manual retrieval from patient samples. The incidence of neuromuscular blockade monitoring was 67% for a selected sample of 200 patients receiving muscle relaxants in the first assessment in 2005. After quality assurance specific guidelines were issued, this percentage increased to 94% 6 months later in another sample of 200 patients. This percentage remained stable in further analyses. The rate of reversal remained steady at around 53%. It should be noted that access to the anaesthesia file PDF database is much easier in comparison to that of the real anaesthesia chart selection and review of archives. 3.2. Pain scores and opioid consumption The percentage of patients entering the PACU with a high pain score (on a five-point scale) despite anticipation is displayed in Fig. 1. To decrease this incidence of high pain scores, we issued a new postoperative analgesic protocol in 2009. The complete results have not yet been extracted because of a database change/upgrade in 2011. Nevertheless, a steady decline in overall morphine consumption has been recorded (Fig. 2). On the contrary, a significant increase in remifentanil (23% in 2002 versus 92% in 2011, P < 0.01) was noted at the expense of sufentanil without affecting overall sedation scores (Fig. 3), thus highlighting an increase in the concomitant use of other non-opioid based analgesics (Fig. 4). 3.3. Postoperative nausea and vomiting (PONV) The percentages associated with the various PONV scores during the study period are represented in Fig. 5. A significant
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
3
100% 90% 80% 70%
4
60%
3
50%
2
40% 30%
1
20%
0
10% 0%
Fig. 1. Pain scores upon arrival in the PACU from 2002 to 2011. Pain was measured via a verbal numerical scale (0 = no pain, 1 = moderate pain, 2 = mild pain, 3 = severe pain, 4 = worst pain imaginable). P < 0.05: year 2011 and 2010 versus 2008.
Dose traon mg/paent 16.00
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 3. Sedation scores on a 5-grade scale (0 = alert; 1 = occasionally drowsy; 2 = asleep, easy to awaken; 3 = difficult to awaken; 4 = unresponsive). No statistical significant difference was found across years.
but mild hypothermia continued to persist in further analyses. A surge was even recorded in 2008 (P < 0.05). However, this occurred simultaneously with a relocation and expansion of operating facilities and recovery rooms (Fig. 6). 4. Discussion
14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
2002
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 2. Postoperative morphine (means SD) consumption for 10 years. These data concern all patients in whom morphine was used in the operating room and PACU. P < 0.05 for years 2011and 2010 versus year 2007.
decline in these scores has occurred since 2006, which was contemporary to the instauration of a new protocol according to standard guidelines and adapted to our cancer patients. The completeness of files as concerns PONV data during the early years were the lowest among the studied indicators. 3.4. Postoperative hypothermia Upon arrival in the PACU, cases where central body temperature was below 35.5 8C were detected. After analysing data for 2005 and 2006 and issuing specific guidelines to decrease hypothermia, severe hypothermia was significantly decreased
This study shows that AIMS could be a useful tool in an anaesthesia department quality assurance program. A key issue in the development of such a program is to carefully select indicators. For certain analyses, a simple keyword search for a specific indicator was sufficient for rapid evaluation and adjustment (e.g. nausea and vomiting, postoperative pain scores, morphine consumption, postoperative hypothermia), while in other situations, such as muscle relaxant monitoring, manual searches and cumbersome evaluations are still necessary. This system did not provide high quality data as concerns muscle relaxant management monitoring, but possible improvements were suggested to the industry in order to obtain more reliable data [9]. While our AIMS was useful in this part of our quality assurance program [10– 12], we recognize that AIMS alone did not permit quality improvement without effective and sustained quality management actions. One should note that a keyword-based search (SQL) required enhanced training of up to several days since our system was not user friendly and intuitive. In order to broaden our use, we changed and upgraded our system in 2011–2012, which provided an increased range of data search indicators and clinical endpoints. However, data extractions required an equivalent amount of training and the new system is still not user friendly. We presented our results to other colleagues, including surgeons, and suggested improvements to a series of measures for each indicator taken into account.
100 90 80 70
KETOPROFEN
60
KETAMINE
50
TRAMADOL
40
PARACETAMOL
30
NEFOPAM
20 10 0
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 4. Trends in the percentages of patients using non-opioid analgesics. A significant rise was noted after a new postoperative analgesic protocol was started in 2009 and continued through 2011 for nefopam, ketoprofen and ketamine (P < 0.05).
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 4
C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
Fig. 5. Postoperative nausea and vomiting scores and related missing files (1 = no nausea and vomiting, 5 = maximum nausea and vomiting). The last columns represent the percentages of missing data, which were high in the initial years. P < 0.05 for PONV scores in year 2007 through 2011 versus year 2006.
File retrieval for certain patients was difficult because system storage is based on the voluntary input of patient names, dates, and procedures. It follows that not all files might be adequately stored, especially those concerning emergency procedures for which human resources are less available. Manual data entry is probably less accurate during vital emergency cases, which raises a legitimate exhaustivity issue. Although accurate data collection remains a primary tool in an anaesthesia quality assurance program, information feedback to other operating room teams is another, crucial component. We regularly proceeded with results feedback and updates to different personnel categories. Nevertheless, we have observed that this sensitization might take several years to be effective in certain areas, such as postoperative hypothermia [13] or PONV. In parallel, we also extracted further indicators, such as anaesthetics drugs or inhaled gas consumption, which served only as economic indicators. We are aware that our approach to quality assurance in anaesthesia was mostly a problem-oriented approach, but this method is as effective as other methods [14]. In addition, effective quality control requires anaesthesia departments to not only develop reliable data extraction and interpretation, but also to provide systems and processes of effective feedback and use of the data for quality improvement [15]. We recognize that we did not use all possible queries that have been proven effective in quality assurance programs (for example, the detection of hypotension). However, these indicators were not directly accessible and required manual extraction. SQL could
indirectly help by finding patients in which vasoconstrictors were used. Nevertheless, this would not be exhaustive because of the possible subjective definition of hypotension. This is also valid for the detection of long-term data for comparing outcomes. Performance analysis or learning curves were not studied. Since the anaesthetic literation addresses cumulative sum analysis (CUSUM) for individual performance [16], we speculate that the database could serve to detect this indicator much more quickly in a broad based population, although this remains to be validated [17]. Nevertheless, individual performance analysis was not our primary objective and CUSUM analysis is still lacking in investigations similar to ours in the anaesthesiology quality assurance program literature. A quality assurance program is almost mandatory in an anaesthesiology department [18], as in all specialties. However, quality excellence in anaesthesia cannot be obtained quickly and requires dedicated system resources and sustained actions. Our specific use of a database for our quality assurance program cannot be generalized because of the specificity of our institution specialized in surgical cancer patients. Funding Funding was from the department of anaesthesia’s own internal funds. Disclosure of interest The authors declare that they have no competing interest. References
Fig. 6. Postoperative hypothermia: a quality assurance program was initiated in 2006 with good results in 2007 (P < 0.05). However, the number of patients with hypothermia less than 35.5 8C increased significantly in 2008 (P < 0.05) due to a new relocated operating room and PACUs with a significant expanded technical platform.
[1] Edsall DW. Quality assessment with a computerized anesthesia information management system (AIMS). QRB Qual Rev Bull 1991;17:182–93. [2] Gregorini P, Manico G, Depatre P. [Computerized recording of physiologic parameters monitored during anesthesia]. Minerva Anestesiol 1993;59: 493–8. [3] Benson M, Junger A, Michel A, Sciuk G, Quinzio L, Marquardt K, et al. Comparison of manual and automated documentation of adverse events with an Anesthesia Information Management System (AIMS). Stud Health Technol Inform 2000;77:925–9. [4] Williams JR. Anesthesia information management systems. AANA J 2005;73:178–81. [5] Simpao AF, Pruitt EY, Cook-Sather SD, Gurnaney HG, Rehman MA. The reliability of manual reporting of clinical events in an anesthesia information management system (AIMS). J Clin Monitor Comput 2012;26:437–9. [6] Benson M, Junger A, Quinzio L, Jost A, Hempelmann G. [Statistics and evaluation of a graphic SQL user interface in an anesthesia information management system (AIMS)]. Anasthesiol Intensivmed Notfallmed Schmerzther 1999;34: 24–32. [7] Motamed C, Bourgain JL. Benefits and possible improvements of an anaesthesia information management system in a quality assurance programme. Br J Anaesth 2014;113:885–6.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx [8] Motamed C, Bourgain JL. Postoperative nausea and vomiting in the postanesthetic care unit, a 5-year survey of a quality assurance program in surgical cancer patients. Bull Cancer 2015;102:405–10. [9] Motamed C, Bourgain JL, D’Hollander A. Survey of muscle relaxant effects management with a kinemyographic-based data archiving system: a retrospective quantitative and contextual quality control approach. J Clin Monitor Comput 2013;27:669–76. [10] Motamed C, Bourgain JL. Impact d’une de´marche d’e´valuation des pratiques professionneles sur l’utilisation du monitorage de la curarisation et l’antoganisation des curares. Ann Fr Anesth Reanim 2009;28:297–301. [11] Motamed C, Bourgain JL. Trend of analgesic consumption and pain scores in the post-anesthetic care unit (a 9-year survey in surgical cancer patients). Bull Cancer 2011;13 [Epub ahead of print]. [12] Motamed C, Salazar G, Bourgain JL. Incidence of severe postoperative pain after cancer surgery despite intraoperative anticipation: a case controlled study. Bull Cancer 2010;97:E37–41.
5
[13] Motamed C, Bourgain JL. Incidence and distribution of inadvertant severe intraoperative hypothermia during cancer surgery: a retrospective single center study. Thrita Journal 2014;3(1):E14744. [14] Balust J, Macario A. Can anesthesia information management systems improve quality in the surgical suite? Curr Opin Anaesthesiol 2009;22:215–22. [15] Benn J, Arnold G, Wei I, Riley C, Aleva F. Using quality indicators in anaesthesia: feeding back data to improve care. Br J Anaesth 2012;109:80–91. [16] Biau DJ, Resche-Rigon M, Godiris-Petit G, Nizard RS, Porcher R. Quality control of surgical and interventional procedures: a review of the CUSUM. Qual Saf Health Care 2007;16:203–7. [17] Starkie T, Drake EJ. Assessment of procedural skills training and performance in anesthesia using cumulative sum analysis (cusum). Can J Anaesth 2013;60:1228–39. [18] Dahmen KG, Albrecht DM. An approach to quality management in anaesthesia: a focus on perioperative care and outcome. Eur J Anaesthesiol Suppl 2001;23:4–9.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
ACCPM-104; No. of Pages 5 Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
Original Article
An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience Cyrus Motamed *, Jean Louis Bourgain Department of Anaesthesia, Gustave-Roussy Cancer Institute, 114, rue E´douard-Vaillant, 94805 Villejuif cedex, France
A R T I C L E I N F O
A B S T R A C T
Article history: Available online xxx
Introduction: Anaesthesia Information Management Systems (AIMS) generate large amounts of data, which might be useful for quality assurance programs. This study was designed to highlight the multiple contributions of our AIMS system in extracting quality indicators over a period of 10 years. Methods: The study was conducted from 2002 to 2011. Two methods were used to extract anaesthesia indicators: the manual extraction of individual files for monitoring neuromuscular relaxation and structured query language (SQL) extraction for other indicators which were postoperative nausea and vomiting (PONV), pain, sedation scores, pain-related medications, scores and postoperative hypothermia. For each indicator, a program of information/meetings and adaptation/suggestions for operating room and PACU personnel was initiated to improve quality assurance, while data were extracted each year. Results: The study included 77,573 patients. The mean overall completeness of data for the initial years ranged from 55 to 85% and was indicator-dependent, which then improved to 95% completeness for the last 5 years. The incidence of neuromuscular monitoring was initially 67% and then increased to 95% (P < 0.05). The rate of pharmacological reversal remained around 53% throughout the study. Regarding SQL data, an improvement of severe postoperative pain and PONV scores was observed throughout the study, while mild postoperative hypothermia remained a challenge, despite efforts for improvement. Discussion: The AIMS system permitted the follow-up of certain indicators through manual sampling and many more via SQL extraction in a sustained and non-time-consuming way across years. However, it requires competent and especially dedicated resources to handle the database. ß 2016 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Keywords: AIMS Quality assurance in anaesthesia Indicators of quality Assurance
1. Introduction The computerized recording of anaesthesia charts, or Anaesthesia Information Management Systems (AIMS), has been available for anaesthesia providers for over two decades [1]. They were originally designed to replace traditional, voluntary, hand recording systems [2]. Automated anaesthesia records and computerized data are described as being more accurate since the phenomena of smoothing, or resistance to the reporting of extreme values, cannot be present [3,4] and, at the least, they create more time for patient-surveillance. Both systems lack total accuracy. For example, the AIMS system does not rely solely on automated parameter recording since self-reporting (e.g.: adverse events) remains necessary and might be under-reported or less accurate [3,5]. In parallel, this information technology generates * Corresponding author. E-mail address: [email protected] (C. Motamed).
huge databases, which allow anaesthesia providers to rapidly consult each anaesthesia record, thus avoiding the waiting time associated with paper record retrieval and transfer. However, data extraction by SQL (Structured Query Language) is also available and can be used to monitor certain data categories, such as intra- or postoperative immediate outcomes [6]. The effective monitoring of medical activity is an integral part of any assurance quality program, together with sustained control of multiple indicators. We previously published a brief report on the overall trends in the main indicators of our quality assurance program for 9 consecutive years [7]. The primary objective of the present study was to quantify the effect of several interventions or changes in practice during 10 consecutive years in our anaesthesia department. 2. Patients and methods In March 2001, our operating rooms were equipped with the ARKTM system integrated with ADU S/5 workstations
http://dx.doi.org/10.1016/j.accpm.2015.11.002 2352-5568/ß 2016 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 2
C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
(Datex-Ohmeda General Electric) for the computerized recording of anaesthesia records and recognized as an AIMS. In order to allow for a washout period during which staff (including users and administrators) became collectively familiar with the system, the actual study period ran from 2002 to 2011. While interventions by anaesthesia providers are indexed manually, physiologic data and vital signs are recorded automatically (e.g.: EKG, pulse, blood pressure, temperature, ventilatory parameters related to anaesthesia machines and inhalational gas consumption). All patients who had anaesthesia, be it general, sedation, regional or local with anaesthesia surveillance, were considered. Upon arrival in the Postoperative Anaesthesia Care Unit (PACU), several clinical items must be manually recorded into the system (e.g.: respiratory status, pain and postoperative nausea and vomiting [PONV], sedation scores and their respective related medications, or central temperature). Our institution’s local ethics committee authorized our department to extract information from the database for our quality assurance program and subsequently publish general results (Avis no 92012/33465). Our clinical procedures strongly encourage anaesthesia providers to index all non-automated administrative, clinical and pharmacological interventions in order to decrease the percentage of incomplete files. During this period, our quality management team started concomitantly monitoring anaesthesia care indicators via manual searches using the PDF storage database (personal anaesthesia file for each patient, Archive Browser1) and by SQL extractions over the entire database for specific indicators. All items extracted by SQL analysis were manually verified for a small number of patients during a short period of time in order to rule out duplicate or missing files. For the purposes of quality assurance in our anaesthesia department, we selected several indicators for extraction: the percentage of pharmacological reversals and instrumental monitoring of neuromuscular blockades, opioid consumption for the intraoperative period and while in the post-anaesthetic care unit (PACU). Other extracted indicators included: pain scores upon arrival, morphine and other analgesic drug consumption, sedation scores, central temperature (tympanic) and PONV scores, which are all manually indexed by the PACU personnel. Results were analysed and meetings on morbidity/mortality were organized in order to issue suggestions for improvements. Indicators were therefore re-checked in further assessments. The suggested interventions and changes in practice were: muscle relaxants: the mandatory use of our neuromuscular monitoring system NMT Datex-Ohmeda1, which is connected to our anaesthesia station and the AIMS system and reversal of neuromuscular blockade and extubation according to updated general clinical guidelines; intra- and postoperative pain management was initially guided by promoting multimodal non-opioid analgesia, and the decrease of postoperative morphine, epidural analgesia using local anaesthetics for major thoracoabdominal surgery, and the progressive shift to use of remifentanil for early extubation; postoperative nausea and vomiting was initially managed at the discretion of the anaesthesiologists. However, after a primary evaluation in 2005 [8], a new protocol using intraoperative dexamethasone and droperidol with odansetron in the PACU was initiated in 2006. This protocol was then readjusted in 2012; postoperative inadvertent hypothermia was discussed every 2 years not only with anaesthesia providers, but also with other operating room actors. Guidelines were issued concerning the management of room temperature, monitoring of intraoperative temperature, and blankets to be available for the patients before entering the operating room area, and the availability of an adequate number of active warming devices.
All data for one patient are recorded as a PDF file, which is printed and added to his/her hospital file when leaving the PACU. The database is also searchable by keywords for system and quality assurance administrators; however, data extraction is not intuitive and requires several hours of training, which can be considered as a limitation of the system. Items, such as the presence and quality of neuromuscular blockade monitoring data (when SQL extraction was not possible), were extracted manually by sampling through a fixed period and for a limited number of patients (200). 2.1. Statistical analysis Results are presented as percentages or means standard deviations when appropriate. Because of the data distribution and the number of missing files in the initial period only, descriptive numbers with their trends are presented. Comparative statistics were performed for a limited number of patients as concerns muscle relaxant monitoring, or for years in which missing files were less than 10% by using ANOVA or Chi2 tests as appropriate. A P value of 0.05 was considered to be statistically significant. 3. Results The data were collected over 10 consecutive years from 2002 to 2011. The total number of patients registered for this period was 77,573. The mean completeness or exhaustivity of the data ranged from 90% in 2002 to > 95% in 2011. However, this was item-dependent; the worst case was the PONV score, which started as low as 50% and attained 86% in 2011. 3.1. Monitoring neuromuscular blockades Although the percentage of reversal could be rapidly accessed by SQL (51 to 57% across years), the incidence of instrumental monitoring could not be extracted from the database and required manual retrieval from patient samples. The incidence of neuromuscular blockade monitoring was 67% for a selected sample of 200 patients receiving muscle relaxants in the first assessment in 2005. After quality assurance specific guidelines were issued, this percentage increased to 94% 6 months later in another sample of 200 patients. This percentage remained stable in further analyses. The rate of reversal remained steady at around 53%. It should be noted that access to the anaesthesia file PDF database is much easier in comparison to that of the real anaesthesia chart selection and review of archives. 3.2. Pain scores and opioid consumption The percentage of patients entering the PACU with a high pain score (on a five-point scale) despite anticipation is displayed in Fig. 1. To decrease this incidence of high pain scores, we issued a new postoperative analgesic protocol in 2009. The complete results have not yet been extracted because of a database change/upgrade in 2011. Nevertheless, a steady decline in overall morphine consumption has been recorded (Fig. 2). On the contrary, a significant increase in remifentanil (23% in 2002 versus 92% in 2011, P < 0.01) was noted at the expense of sufentanil without affecting overall sedation scores (Fig. 3), thus highlighting an increase in the concomitant use of other non-opioid based analgesics (Fig. 4). 3.3. Postoperative nausea and vomiting (PONV) The percentages associated with the various PONV scores during the study period are represented in Fig. 5. A significant
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
3
100% 90% 80% 70%
4
60%
3
50%
2
40% 30%
1
20%
0
10% 0%
Fig. 1. Pain scores upon arrival in the PACU from 2002 to 2011. Pain was measured via a verbal numerical scale (0 = no pain, 1 = moderate pain, 2 = mild pain, 3 = severe pain, 4 = worst pain imaginable). P < 0.05: year 2011 and 2010 versus 2008.
Dose traon mg/paent 16.00
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 3. Sedation scores on a 5-grade scale (0 = alert; 1 = occasionally drowsy; 2 = asleep, easy to awaken; 3 = difficult to awaken; 4 = unresponsive). No statistical significant difference was found across years.
but mild hypothermia continued to persist in further analyses. A surge was even recorded in 2008 (P < 0.05). However, this occurred simultaneously with a relocation and expansion of operating facilities and recovery rooms (Fig. 6). 4. Discussion
14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
2002
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 2. Postoperative morphine (means SD) consumption for 10 years. These data concern all patients in whom morphine was used in the operating room and PACU. P < 0.05 for years 2011and 2010 versus year 2007.
decline in these scores has occurred since 2006, which was contemporary to the instauration of a new protocol according to standard guidelines and adapted to our cancer patients. The completeness of files as concerns PONV data during the early years were the lowest among the studied indicators. 3.4. Postoperative hypothermia Upon arrival in the PACU, cases where central body temperature was below 35.5 8C were detected. After analysing data for 2005 and 2006 and issuing specific guidelines to decrease hypothermia, severe hypothermia was significantly decreased
This study shows that AIMS could be a useful tool in an anaesthesia department quality assurance program. A key issue in the development of such a program is to carefully select indicators. For certain analyses, a simple keyword search for a specific indicator was sufficient for rapid evaluation and adjustment (e.g. nausea and vomiting, postoperative pain scores, morphine consumption, postoperative hypothermia), while in other situations, such as muscle relaxant monitoring, manual searches and cumbersome evaluations are still necessary. This system did not provide high quality data as concerns muscle relaxant management monitoring, but possible improvements were suggested to the industry in order to obtain more reliable data [9]. While our AIMS was useful in this part of our quality assurance program [10– 12], we recognize that AIMS alone did not permit quality improvement without effective and sustained quality management actions. One should note that a keyword-based search (SQL) required enhanced training of up to several days since our system was not user friendly and intuitive. In order to broaden our use, we changed and upgraded our system in 2011–2012, which provided an increased range of data search indicators and clinical endpoints. However, data extractions required an equivalent amount of training and the new system is still not user friendly. We presented our results to other colleagues, including surgeons, and suggested improvements to a series of measures for each indicator taken into account.
100 90 80 70
KETOPROFEN
60
KETAMINE
50
TRAMADOL
40
PARACETAMOL
30
NEFOPAM
20 10 0
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Fig. 4. Trends in the percentages of patients using non-opioid analgesics. A significant rise was noted after a new postoperative analgesic protocol was started in 2009 and continued through 2011 for nefopam, ketoprofen and ketamine (P < 0.05).
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 4
C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx
Fig. 5. Postoperative nausea and vomiting scores and related missing files (1 = no nausea and vomiting, 5 = maximum nausea and vomiting). The last columns represent the percentages of missing data, which were high in the initial years. P < 0.05 for PONV scores in year 2007 through 2011 versus year 2006.
File retrieval for certain patients was difficult because system storage is based on the voluntary input of patient names, dates, and procedures. It follows that not all files might be adequately stored, especially those concerning emergency procedures for which human resources are less available. Manual data entry is probably less accurate during vital emergency cases, which raises a legitimate exhaustivity issue. Although accurate data collection remains a primary tool in an anaesthesia quality assurance program, information feedback to other operating room teams is another, crucial component. We regularly proceeded with results feedback and updates to different personnel categories. Nevertheless, we have observed that this sensitization might take several years to be effective in certain areas, such as postoperative hypothermia [13] or PONV. In parallel, we also extracted further indicators, such as anaesthetics drugs or inhaled gas consumption, which served only as economic indicators. We are aware that our approach to quality assurance in anaesthesia was mostly a problem-oriented approach, but this method is as effective as other methods [14]. In addition, effective quality control requires anaesthesia departments to not only develop reliable data extraction and interpretation, but also to provide systems and processes of effective feedback and use of the data for quality improvement [15]. We recognize that we did not use all possible queries that have been proven effective in quality assurance programs (for example, the detection of hypotension). However, these indicators were not directly accessible and required manual extraction. SQL could
indirectly help by finding patients in which vasoconstrictors were used. Nevertheless, this would not be exhaustive because of the possible subjective definition of hypotension. This is also valid for the detection of long-term data for comparing outcomes. Performance analysis or learning curves were not studied. Since the anaesthetic literation addresses cumulative sum analysis (CUSUM) for individual performance [16], we speculate that the database could serve to detect this indicator much more quickly in a broad based population, although this remains to be validated [17]. Nevertheless, individual performance analysis was not our primary objective and CUSUM analysis is still lacking in investigations similar to ours in the anaesthesiology quality assurance program literature. A quality assurance program is almost mandatory in an anaesthesiology department [18], as in all specialties. However, quality excellence in anaesthesia cannot be obtained quickly and requires dedicated system resources and sustained actions. Our specific use of a database for our quality assurance program cannot be generalized because of the specificity of our institution specialized in surgical cancer patients. Funding Funding was from the department of anaesthesia’s own internal funds. Disclosure of interest The authors declare that they have no competing interest. References
Fig. 6. Postoperative hypothermia: a quality assurance program was initiated in 2006 with good results in 2007 (P < 0.05). However, the number of patients with hypothermia less than 35.5 8C increased significantly in 2008 (P < 0.05) due to a new relocated operating room and PACUs with a significant expanded technical platform.
[1] Edsall DW. Quality assessment with a computerized anesthesia information management system (AIMS). QRB Qual Rev Bull 1991;17:182–93. [2] Gregorini P, Manico G, Depatre P. [Computerized recording of physiologic parameters monitored during anesthesia]. Minerva Anestesiol 1993;59: 493–8. [3] Benson M, Junger A, Michel A, Sciuk G, Quinzio L, Marquardt K, et al. Comparison of manual and automated documentation of adverse events with an Anesthesia Information Management System (AIMS). Stud Health Technol Inform 2000;77:925–9. [4] Williams JR. Anesthesia information management systems. AANA J 2005;73:178–81. [5] Simpao AF, Pruitt EY, Cook-Sather SD, Gurnaney HG, Rehman MA. The reliability of manual reporting of clinical events in an anesthesia information management system (AIMS). J Clin Monitor Comput 2012;26:437–9. [6] Benson M, Junger A, Quinzio L, Jost A, Hempelmann G. [Statistics and evaluation of a graphic SQL user interface in an anesthesia information management system (AIMS)]. Anasthesiol Intensivmed Notfallmed Schmerzther 1999;34: 24–32. [7] Motamed C, Bourgain JL. Benefits and possible improvements of an anaesthesia information management system in a quality assurance programme. Br J Anaesth 2014;113:885–6.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002
G Model
ACCPM-104; No. of Pages 5 C. Motamed, J.L. Bourgain / Anaesth Crit Care Pain Med xxx (2016) xxx–xxx [8] Motamed C, Bourgain JL. Postoperative nausea and vomiting in the postanesthetic care unit, a 5-year survey of a quality assurance program in surgical cancer patients. Bull Cancer 2015;102:405–10. [9] Motamed C, Bourgain JL, D’Hollander A. Survey of muscle relaxant effects management with a kinemyographic-based data archiving system: a retrospective quantitative and contextual quality control approach. J Clin Monitor Comput 2013;27:669–76. [10] Motamed C, Bourgain JL. Impact d’une de´marche d’e´valuation des pratiques professionneles sur l’utilisation du monitorage de la curarisation et l’antoganisation des curares. Ann Fr Anesth Reanim 2009;28:297–301. [11] Motamed C, Bourgain JL. Trend of analgesic consumption and pain scores in the post-anesthetic care unit (a 9-year survey in surgical cancer patients). Bull Cancer 2011;13 [Epub ahead of print]. [12] Motamed C, Salazar G, Bourgain JL. Incidence of severe postoperative pain after cancer surgery despite intraoperative anticipation: a case controlled study. Bull Cancer 2010;97:E37–41.
5
[13] Motamed C, Bourgain JL. Incidence and distribution of inadvertant severe intraoperative hypothermia during cancer surgery: a retrospective single center study. Thrita Journal 2014;3(1):E14744. [14] Balust J, Macario A. Can anesthesia information management systems improve quality in the surgical suite? Curr Opin Anaesthesiol 2009;22:215–22. [15] Benn J, Arnold G, Wei I, Riley C, Aleva F. Using quality indicators in anaesthesia: feeding back data to improve care. Br J Anaesth 2012;109:80–91. [16] Biau DJ, Resche-Rigon M, Godiris-Petit G, Nizard RS, Porcher R. Quality control of surgical and interventional procedures: a review of the CUSUM. Qual Saf Health Care 2007;16:203–7. [17] Starkie T, Drake EJ. Assessment of procedural skills training and performance in anesthesia using cumulative sum analysis (cusum). Can J Anaesth 2013;60:1228–39. [18] Dahmen KG, Albrecht DM. An approach to quality management in anaesthesia: a focus on perioperative care and outcome. Eur J Anaesthesiol Suppl 2001;23:4–9.
Please cite this article in press as: Motamed C, Bourgain JL. An anaesthesia information management system as a tool for a quality assurance program: 10 years of experience. Anaesth Crit Care Pain Med (2016), http://dx.doi.org/10.1016/j.accpm.2015.11.002