Forewarning of hypotensive events using a Bayesian artificial neural network in neurocritical care.

Traumatically brain injured (TBI) patients are at risk from secondary insults. Arterial hypotension, critically low blood pressure, is one of the most dangerous secondary insults and is related to poor outcome in patients. The overall aim of this study was to get proof of the concept that advanced statistical techniques (machine learning) are methods that are able to provide early warning of impending hypotensive events before they occur during neuro-critical care. A Bayesian artificial neural network (BANN) model predicting episodes of hypotension was developed using data from 104 patients selected from the BrainIT multi-center database. Arterial hypotension events were recorded and defined using the Edinburgh University Secondary Insult Grades (EUSIG) physiological adverse event scoring system. The BANN was trained on a random selection of 50% of the available patients (n = 52) and validated on the remaining cohort. A multi-center prospective pilot study (Phase 1, n = 30) was then conducted with the system running live in the clinical environment, followed by a second validation pilot study (Phase 2, n = 49). From these prospectively collected data, a final evaluation study was done on 69 of these patients with 10 patients excluded from the Phase 2 study because of insufficient or invalid data. Each data collection phase was a prospective non-interventional observational study conducted in a live clinical setting to test the data collection systems and the model performance. No prediction information was available to the clinical teams during a patient's stay in the ICU. The final cohort (n = 69), using a decision threshold of 0.4, and including false positive checks, gave a sensitivity of 39.3% (95% CI 32.9-46.1) and a specificity of 91.5% (95% CI 89.0-93.7). Using a decision threshold of 0.3, and false positive correction, gave a sensitivity of 46.6% (95% CI 40.1-53.2) and specificity of 85.6% (95% CI 82.3-88.8). With a decision threshold of 0.3, > 15 min warning of patient instability can be achieved. We have shown, using advanced machine learning techniques running in a live neuro-critical care environment, that it would be possible to give neurointensive teams early warning of potential hypotensive events before they emerge, allowing closer monitoring and earlier clinical assessment in an attempt to prevent the onset of hypotension. The multi-centre clinical infrastructure developed to support the clinical studies provides a solid base for further collaborative research on data quality, false positive correction and the display of early warning data in a clinical setting.

In-house development of a dedicated data acquisition and monitoring system for intracranial pressure, patient posture and patient symptoms in a regional neurosciences centre.

Management of traumatic brain injury and cerebrospinal fluid (CSF) flow disorders can be aided by measurement and monitoring of intracranial pressure (ICP). In addition to pressure measurement, knowledge of patient symptoms and posture during monitoring are also valuable, particularly in the management of CSF flow disorders. ICP monitoring systems have been developed in this centre to meet clinical needs in the absence of commercially available solutions. An early system (mark I) was developed and the technical challenges in its design are described, along with limitations to this system that motivated the development of a new mark II system. The mark II system is then described.

Dental emergencies presenting to a general hospital emergency department in Hobart, Australia.

BACKGROUND: Dental problems presenting to general hospital emergency departments (GHEDs) may create difficulties as there may not be appropriate facilities or personnel to manage them. The incidence and demographics of dental presentations to GHEDs in Tasmania are currently unexamined. This knowledge may be relevant in shaping dental service delivery. METHODS: The Emergency Department Information System (EDIS) at the Royal Hobart Hospital (RHH) was used to audit dental presentations through 2012. The presentations were analysed with respect to demographics and management. A literature review regarding dental presentations to GHEDs was also undertaken. RESULTS: Four hundred and fifty-four patient presentations to the RHH Emergency Department (ED) in 2012 were primarily of a dental nature, comprising 0.91% of all ED presentations. Sixty per cent of patients were male, and most were younger than 30 years. Dental abscesses and toothache accounted for 66%. Nine per cent of dental presentations (including 21% of infections) resulted in hospital admission. CONCLUSIONS: Dental infections contribute a significant burden of ED dental presentations. Encouraging and facilitating patients to seek timely and appropriate primary care from dentists may decrease the number of presentations to ED, and the drain on time and resources. Additionally, this may spare patients from the unpleasant experience of an acute dental infection episode.

Early warning of EUSIG-defined hypotensive events using a Bayesian Artificial Neural Network.

BACKGROUND: Hypotension is recognized as a potentially damaging secondary insult after traumatic brain injury. Systems to give clinical teams some early warning of likely hypotensive instability could be added to the range of existing techniques used in the management of this group of patients. By using the Edinburgh University Secondary Insult Grades (EUSIG) definitions for -hypotension (systolic arterial pressure <90 mmHg OR mean arterial -pressure <70 mmHg) we collected a group of ∼2,000 events by analyzing the Brain-IT database. We then constructed a Bayesian Artificial Neural Network (an advanced statistical modeling technique) that is able to provide some early warning when trained on this previously collected demographic and physiological data. MATERIALS AND METHODS: Using EUSIG defined event data from the Brain-IT database, we identified a Bayesian artificial neural network (BANN) topology and constructed a series of datasets using a group of clinically guided input variables. This allowed us to train a BANN, which was then tested on an unseen set of patients from the Brain-IT database. The initial tests used a particularly harsh assessment criterion whereby a true positive prediction was only allowed if the BANN predicted an upcoming event to the exact minute. We have now developed the system to the point where it is about to be used in a two-stage Phase II clinical trial and we are also researching a more realistic assessment technique. KEY RESULTS: We have constructed a BANN that is able to provide early warning to the clinicians based on a model that uses information from the physiological inputs; systolic and mean arterial pressure and heart rate; and demographic variables age and gender. We use 15-min SubWindows starting at 15 and 30 min before an event and process mean, slope and standard deviations. Based on 10 simulation runs, our current sensitivity is 36.25% (SE 1.31) with a specificity of 90.82% (SE 0.85). Initial results from a Phase I clinical study shows a model sensitivity of 40.95% (SE 6%) and specificity of 86.46% (SE 3%) Although this figure is low it is considered clinically useful for this dangerous condition, provided the false positive rate can be kept sufficiently low as to be practical in an intensive care environment. CONCLUSION: We have shown that using advanced statistical modeling techniques can provide clinical teams with useful information that will assist clinical care.

Trigger characteristics of EUSIG-defined hypotensive events.

BACKGROUND: Hypotension is a recognized -secondary insult after traumatic brain injury (TBI). There are many definitions of hypotension, an often cited example being the Brain Trauma Foundation's current (2007) "Guidelines for the Management of Severe Traumatic Brain Injury," which defines hypotension as systolic pressure <90 mmHg. However, this same document declares "The importance of mean arterial pressure, as opposed to systolic pressure should also be stressed, …." Our work shows that when using the Edinburgh University Secondary Insult Grades (EUSIG) definitions, which require monitoring of both systolic and mean arterial pressures, that most hypotensive events are in fact triggered by a breach of the mean arterial level of 70 mmHg. We suggest that close monitoring of mean arterial pressure would enable clinical teams to avoid more potentially damaging hypotensive events. MATERIALS AND METHODS: An analysis of 100 patients from the Brain-IT database was performed. Using the EUSIG definitions, 2,081 events can be obtained by analyzing the systolic and mean blood pressures on a minute by minute basis. A software program was written to identify and classify the trigger pattern for each event. A categorical analysis of these triggering patterns has been carried out. KEY RESULTS: Our analysis shows that most events are triggered by a drop in mean arterial pressure. In fact a large number of events (91%) occur where the mean arterial pressure is below the threshold limits whereas the systolic pressure does not cross the 90 mmHg limit at all. CONCLUSION: We suggest that more emphasis should be placed on closely monitoring mean arterial pressure as well as systolic pressure when trying to guard against hypotensive problems in traumatically brain injured patients. In future work we will study the underlying physiological mechanisms and attempt to further classify concomitant conditions that may be contributing to the onset of a hypotensive event.

The accuracy of medical history information in referral letters.

BACKGROUND: Accurate medical history information is essential for good patient care and should be notified in the letter of referral. The aim of this study was to investigate the subjective opinion that the medical information in a large number of referrals is either inaccurate or non-existent. METHODS: Medical histories from 54 patients with positive medical history findings upon taking the medical history at the initial consultation appointment were compared to the medical information supplied in the referral letter. RESULTS: Overall, medical information was only 58.8% complete with dental referrals being 55.2% complete and medical referrals 62.4%. The majority of referral letters (70.4%) missed at least one relevant finding and only 29.6% of referrals were 100% complete. CONCLUSIONS: The results of this study suggest that the standard of referral letters needs to be improved as the received referrals were generally incomplete and contained inaccurate information. This highlights the need for each and every practitioner to take their own detailed medical history and not rely on the information supplied in the referral.

Modulating effect of apolipoprotein E polymorphisms on secondary brain insult and outcome after childhood brain trauma.

OBJECTIVE: The aim of this study was to determine the relationship between apolipoprotein E (APO E) alleles, the amount of cerebral perfusion pressure (CPP) insult and outcome in children after brain trauma. MATERIALS AND METHODS: In a prospective two-centre case-control study, the APO E genotypes of 65 critically ill children admitted after brain trauma were correlated with age-related CPP insult quantification, conscious state at the time of discharge from intensive care and global outcome at 6 months post-injury. One hundred sixty healthy age- and sex-matched children were genotyped as controls. RESULTS: The CPP insult level among the e4 carriers with poor outcome was significantly less than the non-e4 carriers (p=0.03). Homozygotic e3 patients with good recovery did so despite having suffered nearly 26 times more CPP insult than those who were not e3 homzygous (p=0.02). CONCLUSION: Different APO E alleles may potentially affect cerebral ischaemic tolerance differently in children after brain trauma.

Temporomandibular joint disorders in patients referred for third molar extraction.

BACKGROUND: Third molar removal has been implicated as a precipitating event for temporomandibular joint disorders. The aim of this study was to determine what proportion of patients had pre-existing pain and dysfunction that could be attributed to the temporomandibular joints. METHODS: Sixty patients referred for third molar removal were clinically examined and a history of their presenting complaint recorded at the initial consultation visit. Patients were then diagnosed and categorized. RESULTS: Of the total number of patients examined, 13.3 per cent showed signs and symptoms of temporomandibular joint pain and dysfunction while a further 23.3 per cent also had symptomatic third molar teeth. CONCLUSIONS: The results of this study suggest that the signs of temporomandibular joint disorders are common in patients referred for third molar extractions.

The use of hyperventilation therapy after traumatic brain injury in Europe: an analysis of the BrainIT database.

OBJECTIVE: To assess the use of hyperventilation and the adherence to Brain Trauma Foundation-Guidelines (BTF-G) after traumatic brain injury (TBI). SETTING: Twenty-two European centers are participating in the BrainIT initiative. DESIGN: Retrospective analysis of monitoring data. PATIENTS AND PARTICIPANTS: One hundred and fifty-one patients with a known time of trauma and at least one recorded arterial blood-gas (ABG) analysis. MEASUREMENTS AND RESULTS: A total number of 7,703 ABGs, representing 2,269 ventilation episodes (VE) were included in the analysis. Related minute-by-minute ICP data were taken from a 30 min time window around each ABG collection. Data are given as mean with standard deviation. (1) Patients without elevated intracranial pressure (ICP) (< 20 mmHg) manifested a statistically significant higher P(a)CO(2) (36 +/- 5.7 mmHg) in comparison to patients with elevated ICP (> or = 20 mmHg; P(a)CO(2): 34 +/- 5.4 mmHg, P < 0.001). (2) Intensified forced hyperventilation (P(a)CO(2) < or = 25 mmHg) in the absence of elevated ICP was found in only 49 VE (2%). (3) Early prophylactic hyperventilation (< 24 h after TBI; P(a)CO(2) < or = 35 mmHg, ICP < 20 mmHg) was used in 1,224 VE (54%). (4) During forced hyperventilation (P(a)CO(2) < or = 30 mmHg), simultaneous monitoring of brain tissue pO(2) or S(jv)O(2) was used in only 204 VE (9%). CONCLUSION: While overall adherence to current BTF-G seems to be the rule, its recommendations on early prophylactic hyperventilation as well as the use of additional cerebral oxygenation monitoring during forced hyperventilation are not followed in this sample of European TBI centers. DESCRIPTOR: Neurotrauma.

Low frequency pressure waves of possible autonomic origin in severely head-injured children.

BACKGROUND: Useful information (both clinical and pathophysiological) which may be extracted from intracranial pressure (ICP) recordings include: (1) the mean level of ICP (and CPP), (2) cerebrovascular autoregulation status, (3) the intracranial pulse pressure (the pulse wave index, ICPpp/ICPm) or the pressure-volume compensatory reserve index (RAP) and (4) the presence of any abnormal ICP waveform. This paper describes a slow frequency ICP waveform in children with TBI and postulates the pathophysiological basis and whether it contains clinically useful detail. METHODS: Children admitted to the Regional Head Injury Service in Edinburgh with TBI have continuously monitored ICP, MAP, CPP, and other physiological data (stored at a 1-min resolution). Slow frequency waveforms were noted, prompting a review of the stored monitoring from all cases over a 10 year period. FINDINGS: Episodic slow pressure waves were detected in 11 of 122 severely head-injured (HI) children. The waveforms were detected in children of all ages (1.6-15 years) in the ICP signal, which were in phase with similar fluctuations in the MAP, CPP, and HR signals. Their mean periodicity was 1 per 7 min (range 1 per 5-10 min), with a mean ICP pulse wave amplitude of 5.45 mmHg (range 4-7.5), and mean MAP pulse wave amplitude (pulse pressure) of 10.4 mmHg (range 4-15 mmHg). The duration was variable (range approx 2 h to 4.5 days). They were detected in the preterminal phase after serious HI, as well as in those children who made an independent recovery (GOS 4/5). The waves were not related to the mean levels of ICP, CPP, MAP, temperature or the state of cerebrovascular autoregulation. CONCLUSIONS: We postulate that these previously unreported slow waveforms may reflect the very low frequency (VLF) and ultra low frequency (ULF; < or = 1 per 5 min) components of heart rate and arterial blood pressure variability.

Are head injury guidelines changing the outcome of head injured children? A regional investigation.

BACKGROUND: Secondary pathophysiological CPP insult is related to outcome after head injury, and improved management would be expected to reduce secondary brain insult. Paediatric head injury management guidelines have been published in recent years, by SIGN (2000), RCPCH (2001), NICE (June 2003), and jointly by Critical/Intensive Care Societies (C/ICS July 2003). We investigated whether outcome of children's head injury (and total burden of secondary CPP insult) has changed (1) annually; (2) before and after the introduction of any HI guidelines, and (3) following other service changes. METHODS: Seventy-six children (aged 1-14 years with severe HI) were admitted to the Edinburgh Regional Head Injury Service between 1989 and 2006, and dichotomised at various time points and compared in terms of: demographic factors, intracranial pressure (ICP), cerebral perfusion pressure (CPP) insults [e.g. age-banded pressure-time index (PTI)], and Glasgow Outcome Scale (GOS) score (assessed at 6 months post injury). FINDINGS: When dichotomised around the SIGN guidelines, there were no statistically significant differences between the two group's demography or in primary brain injury, but the outcomes were different (p = 0.03), with 6 vs 4 GOS1 (died), 2 vs 4 GOS3 (severely disabled), 5 vs 16 GOS4 (moderately disabled) and 23 vs 14 GOS5 (good recovery), when comparing before and after year 2000. GOS4 was significantly different (chi-square = 7.99, p < 0.007). There was a (non-significant) trend for the later years to have longer insult durations of ICP, hypertension, CPP, hypoxia, pyrexia, tachycardia and bradycardia, greater PTI for both CPP and ICP, and more CPP insults (p = 0.003). There was, however, significantly less CPP insult (p = 0.030) after the introduction of the more management-oriented C/ICS guidelines. CONCLUSIONS: The most recent paediatric HI guidelines appear to have reduced the burden of secondary insult, but more time is required to determine if this will be reflected in improved outcomes.

The measurement of brain tissue stiffness in-vivo.

BACKGROUND: There is considerable interest in surgical decompression as a management strategy (RescueICP) for intractable intracranial hypertension. After such an operation measurements of intracranial pressure (ICP) and thus cerebral perfusion pressure (CPP) become less meaningful. Measurements of the biomechanical properties of the brain may be one measure capable of detecting changing status of such patients. However these properties of the brain are neither documented or well understood. We have developed an indentation probe capable of making measurements of human brain stiffness. METHOD: The device consists of an indenting tip of depth 2 mm and diameter 12 mm surrounded by an annular body of 20 mm diameter. Measurements are made by two load cells, connected through interface electronics to a laptop computer. FINDINGS: Laboratory measurements show the probe to provide accurate and repeatable measurements over a range of zero to 10N. Inter-operator variability from six healthcare professionals had a coefficient of variance of 8.75%. Measurements obtained during surgery from a patient undergoing tumour resection were towards the lower end of the device's measurable range. CONCLUSIONS: We have determined that this indentation device has a linear response and that the inter- and intra-operator variability is low. Although the device is still in an early stage of development, preliminary results during intracranial surgery demonstrate that this device is capable of measuring in-vivo tissue stiffness. Further work is required to derive a quantitative "stiffness index" from the two load curves. In addition a standard operation method is required so that consistent and repeatable measurements are made. The device may be of value in assessing patients after decompressive craniectomy.

BrainIT: a trans-national head injury monitoring research network.

BACKGROUND: Studies of therapeutic interventions and management strategies on head injured patients are difficult to undertake. BrainIT provides validated data for analysis available to centers that contribute data to allow post-hoc analysis and hypothesis testing. METHODS: Both physiological and intensive care management data are collected. Patient identification is eliminated prior to transfer of data to a central database in Glasgow. Requests for missing/ ambiguous data are sent back to the local center. Country coordinating centers provide advice, training, and assistance to centers and manage the data validation process. RESULTS: Currently 30 centers participate in the group. Data collection started in January 2004 and 242 patients have been recruited. Data validation tools were developed to ensure data accuracy and all analysis must be undertaken on validated data. CONCLUSION: BrainIT is an open, collaborative network that has been established with primary objectives of i) creating a core data set of information, ii) standardizing the collection methodology, iii) providing data collection tools, iv) creating and populating a data base for future analysis, and v) establishing data validation methodologies. Improved standards for multi-center data collection should permit the more accurate analysis of monitoring and management studies in head injured patients.

Variability in pupil size estimation.

BACKGROUND: The clinical estimation of pupil size and reactivity is central to the neurological assessment of patients, particularly those with or at risk of neurological damage. Health care professionals who examine pupils have differing levels of skill and training, yet their recordings are passed along the patient care pathway and can influence care decisions. The aim of this study was to determine if any statistical differences existed in the estimation of pupil size by different groups of health care professionals. METHODS: A total of 102 health care professionals working in the critical care environment were asked to estimate and record the pupil size of a series of 12 artificial eyes with varying pupil diameter and iris colour. All estimations were performed indoors under ambient lighting conditions. RESULTS: Our results established a statistically significant difference between staff groups in the estimation of pupil size. CONCLUSION: The demonstrated variability in pupil size estimation may not be clinically significant. However, it remains desirable to have consistency of measurement throughout the patient care pathway.

Critical thresholds of intracranial pressure and cerebral perfusion pressure related to age in paediatric head injury.

BACKGROUND: The principal strategy for managing head injury is to reduce the frequency and severity of secondary brain insults from intracranial pressure (ICP) and cerebral perfusion pressure (CPP), and hence improve outcome. Precise critical threshold levels have not been determined in head injured children. OBJECTIVE: To create a novel pressure-time index (PTI) measuring both duration and amplitude of insult, and then employ it to determine critical insult thresholds of ICP and CPP in children. METHODS: Prospective, observational, physiologically based study from Edinburgh and Newcastle, using patient monitored blood pressure, ICP, and CPP time series data. The PTI for ICP and CPP for 81 children, using theoretical values derived from physiological norms, was varied systematically to derive critical insult thresholds which delineate Glasgow outcome scale categories. RESULTS: The PTI for CPP had a very high predictive value for outcome (receiver operating characteristic analyses: area under curve = 0.957 and 0.890 for mortality and favourable outcome, respectively) and was more predictive than for ICP. Initial physiological values most accurately predicted favourable outcome. The CPP critical threshold values determined for children aged 2-6, 7-10, and 11-15 years were 48, 54, and 58 mm Hg. respectively. CONCLUSIONS: The PTI is the first substantive paediatric index of total ICP and CPP following head injury. The insult thresholds generated are identical to age related physiological values. Management guidelines for paediatric head injuries should take account of these CPP thresholds to titrate appropriate pressor therapy.

The clinical application of non-invasive intracranial blood volume pulse wave monitoring.

An ultrasonic method was used to non-invasively measure intracranial blood volume (IBV) pulse waveforms. This technology has previously shown a strong association between invasively recorded ICP pulse waves and non-invasively recorded IBV pulse waves. The objective of the present study was to investigate the diagnostic value of non-invasively measured IBV pulse waves in the cases of different pathologies. A total of 75 patients were examined and these included cases of acute, chronic and stabilized hydrocephalus, spinal cord injury and terminal blood flow. These were compared to a control group of 53 healthy volunteers. The object of comparison was normalized and averaged IBV pulse waves. Pathological IBV pulse waveforms were compared with IBV pulse waveforms of the normal group using sub-wave values, the area under waveform curve and the Euclidean distance calculation. The non-invasively measured IBV pulse waveform is not significantly dependent on acoustic path, gender or age. A detectable change in IBV pulse waveform shape was observed in situations when disturbance in intracranial hydrodynamics was present, e.g. during hypoventilation tests, in cases of terminal blood flow and hydrocephaly, depicting the level of hydrocephalus activity and the patient's compensatory capabilities as well as the effect of treatment.

Clinical comparison of tympanic membrane displacement with invasive intracranial pressure measurements.

A non-invasive method of assessing intracranial pressure (ICP) would be of benefit to patients with abnormal cerebral pathology that could give rise to changes in ICP. In particular, it would assist the regular monitoring of hydrocephalus patients. This study evaluated a technique using tympanic membrane displacement (TMD) measurements, which has been reported to provide a reliable, non-invasive measure of ICP. A group of 135 hydrocephalus patients was studied, as well as 13 patients with benign intracranial hypertension and a control group of 77 volunteers. TMD measurements were carried out using the Marchbanks measurement system analyser and compared between the groups. In 36 patients, invasive measurements of ICP carried out at the same time were compared with the TMD values. A highly significant relationship was found between TMD and ICP but intersubject variability was high and the predictive value of the technique low. Taking the normal range of ICP to be 10-15 mmHg, the predictive limits of the regression are an order of magnitude wider than this and therefore Vm cannot be used as a surrogate for ICP. In conclusion, TMD measurements do not provide a reliable non-invasive measure of ICP in patients with shunted hydrocephalus.

The homeodomain protein Nanog and pluripotency in mouse embryonic stem cells.

Intrinsic regulators of the pluripotency of mouse ES (embryonic stem) cells include the homeodomain proteins Oct4 and the recently identified Nanog. When overexpressed, Nanog displays the unique attribute of robustly sustaining ES cell self-renewal in the absence of the otherwise requisite extracellular stimulation by LIF (leukaemia inhibitory factor) and BMP (bone morphogenetic protein). Here, we review our current understanding of the function of Nanog in pluripotent stem cells both in vitro and in vivo.

A cluster of culture positive gonococcal infections but with false negative cppB gene based PCR.

OBJECTIVES: To describe the prevalence and characteristics of isolates of Neisseria gonorrhoeae grown from urine samples that produced negative results with nucleic acid amplification assays (NAA) targeting the cppB gene. METHODS: An initial cluster of culture positive, but cppB gene based NAA negative, gonococcal infections was recognised. Urine samples and suspensions of gonococci isolated over 9 months in the Northern Territory of Australia were examined using cppB gene based and other non-cppB gene based NAA. The gonococcal isolates were phenotyped by determining the auxotype/serovar (A/S) class and genotyped by pulsed field gel electrophoresis (PFGE). RESULTS: 14 (9.8%) of 143 gonococci isolated were of A/S class Pro(-/)Brpyut, indistinguishable on PFGE and negative in cppB gene based, but not other, NAA. CONCLUSIONS: This cluster represents a temporal and geographic expansion of a gonococcal subtype lacking the cppB gene with consequent loss of sensitivity of NAA dependent on amplification of this target. Gonococci lacking the cppB gene have in the past been more commonly associated with the PAU-/PCU- auxotype, a gonococcal subtype hitherto infrequently encountered in Australia. NAA based on the cppB gene as a target may produce false positive as well as false negative NAA. This suggests that unless there is continuing comparison with culture to show their utility, cppB gene based NAA should be regarded as suboptimal for use either as a diagnostic or supplemental assay for diagnosis of gonorrhoea, and NAA with alternative amplification targets should be substituted.