Development and usage of an anesthesia data warehouse: lessons learnt from a 10-year project.

This paper describes the development and implementation of an anesthesia data warehouse in the Lille University Hospital. We share the lessons learned from a ten-year project and provide guidance for the implementation of such a project. Our clinical data warehouse is mainly fed with data collected by the anesthesia information management system and hospital discharge reports. The data warehouse stores historical and accurate data with an accuracy level of the day for administrative data, and of the second for monitoring data. Datamarts complete the architecture and provide secondary computed data and indicators, in order to execute queries faster and easily. Between 2010 and 2021, 636 784 anesthesia records were integrated for 353 152 patients. We reported the main concerns and barriers during the development of this project and we provided 8 tips to handle them. We have implemented our data warehouse into the OMOP common data model as a complementary downstream data model. The next step of the project will be to disseminate the use of the OMOP data model for anesthesia and critical care, and drive the trend towards federated learning to enhance collaborations and multicenter studies.

Relevance of a novel external dynamic distraction device for treating back pain.

Low back pain is a common, expensive, and disabling condition in industrialized countries. There is still no consensus for its ideal management. Believing in the beneficial effect of traction, we developed a novel external dynamic distraction device. The purpose of this work was to demonstrate that external distraction allows limiting the pressure exerted in standing-up position on the lower intervertebral discs. Numerical and cadaveric studies were used as complementary approaches. Firstly, we implemented the device into a numerical model of a validated musculoskeletal software (Anybody Modeling System) and we calculated the lower disc pressure while traction forces were applied. Secondly, we performed an anatomical study using a non-formalin preserved cadaver placed in a sitting position. A pressure sensor was placed in the lower discs under fluoroscopic control through a Jamshidi needle. The intradiscal pressure was then measured continuously at rest while applying a traction force of 200 N. Both numerical and cadaveric studies demonstrated a decrease in intradiscal pressures after applying a traction force with the external device. Using the numerical model, we showed that tensile forces below 500 N in total were sufficient. The application of higher forces seems useless and potentially deleterious. External dynamic distraction device is able to significantly decrease the intradiscal pressure in a sitting or standing position. However, the therapeutic effects need to be proven using clinical studies.

Short-term Impact of Assisted Deliveries: Evaluation Based on Behavioral Pain Scoring and Heart Rate Variability.

OBJECTIVES: Assisted deliveries (ADs) are used in current practice by obstetrical teams during labor when the fetus is likely to face difficulties. In this study, we hypothesized that pain related to instrumental delivery could impair autonomic nervous system (ANS) activity. MATERIALS AND METHODS: We investigated infants born by spontaneous delivery (SD group, n=35) and those whose deliveries were assisted by vacuum, forceps, or both (AD group, n=35) between the second and fourth hours of life. Pain was evaluated using the Echelle de Douleur et d'Inconfort du Nouveau-Né/Neonatal Pain and Discomfort Scale (EDIN) behavioral pain scale. ANS activity was analyzed using the newborn infant parasympathetic evaluation (NIPE) index, a heart rate variability based index which is related to the autonomic activity mediated by the parasympathetic nervous system. RESULTS: Neonates in the AD group presented higher EDIN scores than neonates born by spontaneous vaginal deliveries (SD group) (P<0.0001). In contrast, the NIPE index was significantly reduced in the AD group compared with the SD group (P=0.005). A significant inverse correlation was found between the NIPE index and the EDIN score (r=-0.287, P=0.016). DISCUSSION: ADs (vacuum, forceps, or both) are associated with persistent pain after birth, unlike normal vaginal deliveries. Moreover ADs are associated with reduced NIPE. Taken together, our results suggest that pain related to instrumental delivery impairs the ANS activity.

Parasympathetic tone variations according to umbilical cord pH at birth: a computerized fetal heart rate variability analysis.

Non-reassuring fetal heart rate tracings reflect an imbalance between the parasympathetic and sympathetic nervous systems. In this situation, fetal asphyxia can be suspected and may be confirmed by metabolic measurements at birth like low pH or high base deficit values. The objective of this study was to determine whether fetal asphyxia during labor is related to parasympathetic nervous system activity. This is a retrospective study of a database collected in 5 centers. Two hundred and ninety-nine fetal heart rate tracings collected during labor were analyzed. Autonomic nervous system, especially the parasympathetic nervous system, was analyzed using an original index: the FSI (Fetal Stress Index). The FSI is a parasympathetic activity evaluation based on fetal heart rate variability analysis. Infants were grouped based on normal or low pH value at birth. FSI was measured during the last 30 min of labor before birth and compared between groups. The minimum value of the FSI during the last 30 min before delivery was significantly lower in the group with the lower umbilical cord arterial pH value. In this pilot study during labor, FSI was lower in the group of infants with low arterial pH at birth.

[Analysing heart rate variability to improve the monitoring of pain]. / Analyser la variabilité de la fréquence cardiaque pour mieux monitorer la douleur.

An innovative technique based on the analysis of instantaneous heart rate variability helps to improve the prevention and management of pain and discomfort. Simple to implement, this non-invasive technique is based on the continuous recording of the electrocardiograph signal.

Assessment of Procedural Pain in Children Using Analgesia Nociception Index: A Pilot Study.

OBJECTIVES: The Analgesia Nociception Index (ANI), based on heart rate (HR) variability analysis, is known to decrease after a painful stimulus during surgery under general anesthesia in adults. It is measured continuously and noninvasively. We studied ANI response to procedural pain in a pediatric population and ANI measurement's feasibility in this context, across age. METHODS: A prospective, noninterventional pilot study was performed. All children (between 6 mo and under 18 y) undergoing muscle biopsy conducted under analgesia and light sedation were included. Medical staff was blind to the ANI monitor. HR and ANI were recorded and analyzed during 2 periods: T1 before incision and T2 after incision. Pain was assessed by the FLACC scale at T2. We observed ANI and HR variations after incision. ANI, HR, and FLACC were compared between children younger or older than 6 years. Enrollment or technical issues were reported. RESULTS: A total of 26 children were included (median age, 6 y; ranging from 6 mo to 16 y; 16 male). ANI decreased from T1 to T2. HR, ANI, or FLACC values were not different in children younger or older than 6 years. No parents or children refused to take part in the study. No technical issues was reported. DISCUSSION: In this pilot study, ANI measurement seems relevant in pediatric procedural pain, across age. Further studies are needed to confirm these results.

Methodology to automatically detect abnormal values of vital parameters in anesthesia time-series: Proposal for an adaptable algorithm.

Abnormal values of vital parameters such as hypotension or tachycardia may occur during anesthesia and may be detected by analyzing time-series data collected during the procedure by the Anesthesia Information Management System. When crossed with other data from the Hospital Information System, abnormal values of vital parameters have been linked with postoperative morbidity and mortality. However, methods for the automatic detection of these events are poorly documented in the literature and differ between studies, making it difficult to reproduce results. In this paper, we propose a methodology for the automatic detection of abnormal values of vital parameters. This methodology uses an algorithm allowing the configuration of threshold values for any vital parameters as well as the management of missing data. Four examples illustrate the application of the algorithm, after which it is applied to three vital signs (heart rate, SpO2, and mean arterial pressure) to all 2014 anesthetic records at our institution.

Ictal changes in parasympathetic tone: Prediction of postictal oxygen desaturation.

OBJECTIVE: To measure changes in parasympathetic tone before, during, and after temporal seizures, and to determine whether changes in high-frequency heart rate variability are correlated with postictal oxygen desaturation. METHODS: We recorded the electrocardiogram and peripheral oxygen saturation during 55 temporal lobe seizures and calculated a high-frequency variability index (HFVI) as a marker of parasympathetic tone for periods of 20 minutes (centered on seizure onset). We then compared HFVI values in seizures with and without postictal hypoxemia, and looked for correlations between HFVI changes and the risk of sudden unexpected death in epilepsy (SUDEP) (as assessed with the SUDEP-7 Inventory). RESULTS: Parasympathetic tone decreased rapidly at the onset of temporal lobe seizures, reached its minimum value at the end of the seizure, and then gradually returned to its preictal value. Changes in parasympathetic tone were more intense and longer-lasting in older patients with a longer duration of epilepsy. The HFVI was significantly lower during seizures with hypoxemia, and remained significantly lower 5 minutes after the end of the seizure. The change in the HFVI slope over the first 30 seconds of the seizure was predictive of postictal oxygen desaturation. Postictal autonomic changes were correlated with the SUDEP-7 scores. CONCLUSION: Our results showed that ictal autonomic dysfunction is correlated with postictal hypoxemia. A prolonged impairment of parasympathetic tone might expose a patient to a greater risk of postictal sudden unexpected death. The real-time measurement of parasympathetic tone in patients with epilepsy may be of value to medical staff as an early warning system.

A substitution method to improve completeness of events documentation in anesthesia records.

AIMS are optimized to find and display data and curves about one specific intervention but is not retrospective analysis on a huge volume of interventions. Such a system present two main limitation; (1) the transactional database architecture, (2) the completeness of documentation. In order to solve the architectural problem, data warehouses were developed to propose architecture suitable for analysis. However, completeness of documentation stays unsolved. In this paper, we describe a method which allows determining of substitution rules in order to detect missing anesthesia events in an anesthesia record. Our method is based on the principle that missing event could be detected using a substitution one defined as the nearest documented event. As an example, we focused on the automatic detection of the start and the end of anesthesia procedure when these events were not documented by the clinicians. We applied our method on a set of records in order to evaluate; (1) the event detection accuracy, (2) the improvement of valid records. For the year 2010-2012, we obtained event detection with a precision of 0.00 (-2.22; 2.00) min for the start of anesthesia and 0.10 (0.00; 0.35) min for the end of anesthesia. On the other hand, we increased by 21.1% the data completeness (from 80.3 to 97.2% of the total database) for the start and the end of anesthesia events. This method seems to be efficient to replace missing "start and end of anesthesia" events. This method could also be used to replace other missing time events in this particular data warehouse as well as in other kind of data warehouses.

Maintenance of parasympathetic inhibition following emotional induction in patients with restrictive type anorexia nervosa.

This study aimed to explore changes in heart rate variability (HRV), a proxy for parasympathetic activity characterizing emotion regulation processes before, during and after negative emotional induction in patients suffering from restrictive type anorexia nervosa (AN-RT). We compared two methods of HRV analysis, the Fast Fourier Transform high frequency (FFT-HF) and a specific HRV high frequency analysis technique, namely, the wavelet transform HRV (WT-HRV). A sample of 16 inpatients with AN-RT was compared to 24 control participants. Heart rate (HR) was continuously recorded for 5 min before the beginning of the video until 5 min after the video. The participants answered questionnaires concerning their eating behaviors, mood disorders and difficulties in emotion regulation. During the entire procedure, the FFT-HF in patients was lower than that in controls. Using the WT-HRV, the patients did not differ from the controls at baseline, and only the controls showed a decrease during emotional induction. After the video, the WT-HRV in patients began to decrease during the first 2 min of emotional recovery although the WT-HRV in controls was already increased. These results highlighted the disturbances in the physiological dynamics of emotion regulation processes in patients with AN-RT.

Variations of the analgesia nociception index during propofol anesthesia for total knee replacement.

OBJECTIVE: The Analgesia Nociception Index (ANI) monitor measures the relative parasympathetic tone as a surrogate for the antinociception/nociception balance during general anesthesia. The aims of this observational study were to test whether ANI could early detect hemodynamic reactivity (HemodR) during propofol anesthesia, measure pain in conscious patients after surgery, and determine ANI predictive thresholds. MATERIALS AND METHODS: After institutional approval and informed consent, adult patients scheduled for total knee replacement were included. Propofol was administered using a target-controlled infusion device, and sufentanil was administered at induction and in case of "HemodR," defined as a 20% increase in heart rate or systolic blood pressure. Data were collected before the start of surgery (NoStim), in case of "HemodR," and after awakening before and after the pain had been treated by truncular analgesia. Nonparametric tests were conducted. Thresholds were determined using an receiver operating characteristic analysis. Results are presented as median (interquartile range). RESULTS: A total of 27 patients have been analysed. ANI decreased from 82 (30) at "NoStim" to 47 (22) at "HemodR," whereas heart rate increased moderately from 61 (14) to 65 (18) and systolic blood pressure increased significantly from 91 mm Hg (16 mm Hg) to 151 mm Hg (25 mm Hg). Receiver operating characteristic analysis led to a threshold of 63 for "HemodR" detection (sensitivity=80%, specificity=88%, area under the curve=0.92), whereas ANI performance in awake patients was lower. DISCUSSION: ANI measures during propofol anesthesia are coherent with the evolution of the analgesia/nociception balance, although its performance decreases in awake patients. Further clinical validation should focus on demonstrating the benefit of maintaining ANI over 63 during surgery.

Newborn infant pain assessment using heart rate variability analysis.

OBJECTIVES: Systems controlling cardiovascular function are closely coupled with the perception of pain. Heart rate variability (HRV) is a well-established noninvasive measure of cardiac autonomic control. We hypothesized that pain may alter HRV in the newborn infant and that HRV analysis could be used as an indicator of prolonged pain in the newborn infant. METHODS: To test the hypothesis, we measured the magnitude of the heart rate high-frequency variations using an innovative High Frequency Variability Index (HFVI) in newborn infants at risk of postoperative pain. We investigated newborn infants with a gestational age (GA) more than 34 weeks, and who were admitted after a major surgical procedure. Inclusions ranged from 2 to 72 hours after the surgery. The postoperative pain was scored using EDIN scale (neonatal pain and discomfort scale) at the end of the 2 hours recording period. The infants were separated in: (1) Group "Low EDIN," when EDIN<5; and (2) Group "High EDIN," when EDIN >=5. Predictive positive and negative values of a threshold value of HFVI in assessing pain have been studied. RESULTS: Twenty-eight newborn infants were enrolled in the study (mean GA=37.8+/-1.5 wk) at a median delay between the surgery and the recording of 5 hours. Mean EDIN were 2+/-1 and 7+/-2 in respectively the groups "Low EDIN" and "High EDIN." The 2 groups were similar for GA, basal heart and respiratory rates, SpO2, mean arterial blood pressure, and morphine infusion rate. HFVI was significantly lower in the group "High EDIN" than in the group "Low EDIN" (0.7+/-0.2 vs. 1.2+/-0.3, respectively; P<0.01). An HFVI <0.9 was able to predict an EDIN score >=5, with a sensitivity of 90%, and a specificity of 75%. DISCUSSION: The results of this study indicate that postoperative pain is associated with a decreased high-frequency HRV in full-term newborn infants. Our findings suggest that HRV could be used as an indicator to assess prolonged pain in the newborn infants.

Heart rate variability during total intravenous anesthesia: effects of nociception and analgesia.

BACKGROUND: Changes in heart rate variability (HRV) during anesthesia likely result from the interaction of hypnosis, surgical stimulation, analgesia and direct cardiovascular effects of drugs, but the interaction between these variables is unclear. This study was designed to characterize the impact of both surgical nociception and analgesia on HRV in propofol-anesthetized patients. METHODS: HRV was analyzed using wavelet transform in 49 patients (ASA status 1-2) before induction of anesthesia and then throughout stable anesthesia with propofol, in the absence of nociceptive stimulation, and then during surgery, in the presence of deep (adequate) or light (inadequate) analgesia provided by various opioids (sufentanil [n=19], alfentanil [n=18], or remifentanil [n=12]. RESULTS: Anesthesia reduced total power as well as high frequency (HF) and low frequency (LF) powers (all: P<0.01), with an increase (P=0.002) in the proportional part of HF power (HFnu). During nociception, HFnu decreased in a sensitive and reproducible way (P<0.01) in case of light analgesia, whereas HRV did not change when patient received adequate analgesia. CONCLUSIONS: The nociception-analgesia balance is a direct determinant of HRV during surgical anesthesia. HFnu may behave like an early indicator of inadequate analgesia. These results have potential implication for monitoring adequacy of analgesia in healthy patients undergoing intravenous anesthesia. Additional work is needed for application across patient populations.

Evaluation of spinal curvatures after a recent osteoporotic vertebral fracture.

OBJECTIVES: To evaluate spinal curvature changes over a 3-year period in postmenopausal women who had had an osteoporotic vertebral fracture within the last 3 months. METHODS: Thoracic kyphosis and lumbar lordosis were measured using a curviscope at baseline and after 1, 3, 6, 12, and 36 months. Anteroposterior and lateral radiographs of the thoracolumbar spine were obtained after 1 and 3 years. RESULTS: Sixty-one patients were included. At baseline, a significant increase in thoracic curvature was found in the subgroup with thoracic fractures as compared to the subgroups with thoracolumbar or lumbar fractures (64 degrees +/- 9 degrees, 56 degrees +/- 10, and 56 degrees +/- 13, respectively; P < 0.05). No lumbar curvature differences were found. Thoracic curvature was significantly correlated with age (r = -0.48, P < 0.001) and with the vertebral deformity index (r = 0.6, P < 0.001). A significant increase in thoracic curvature was apparent 3 months into the study; after 3 years, the increase was 5.6 degrees +/- 0.7 (P < 0.01). A moderate increase in lumbar curvature was found after 3 years (P < 0.01). Five of 13 patients and five of 10 patients had at least one incident fracture after 1 and 3 years, respectively. Mean thoracic curvature was greater among the patients with than without incident fractures after 1 and 3 years, although the difference was not statistically significant. CONCLUSION: Thoracic compression fractures significantly increase thoracic kyphosis as compared to dorsolumbar and lumbar fractures. Thoracic kyphosis worsens overtime in patients with prevalent vertebral fractures. These data invite an evaluation of techniques capable of providing early correction of alignment disorders, such as widespread use of bracing or kyphoplasty.