Dissertations / Theses on the topic 'Anaesthesia and intensive care'

Mathematical models of physiological processes can be used in critical care and anaesthesia to improve the understanding of disease processes and to guide treatment. This thesis provides a detailed description of two studies that are related through their shared aim of modelling different aspects of brain physiology. The Relationship Between Transcranial Bioimpedance and Invasive Intracranial Pressure Measurement in Traumatic Brain Injury Patients (BioTBI) Study describes an attempt to model intracranial pressure (ICP) in patients admitted with severe traumatic brain injury (TBI). It is introduced with a detailed discussion of the monitoring and modelling of ICP in patients with TBI alongside the rationale for considering transcranial bioimpedance (TCB) as a non-invasive approach to estimating ICP. The BioTBI Study confirmed a significant relationship between TCB and invasively measured ICP in ten patients admitted to the neurological intensive care unit (NICU) with severe TBI. Even when using an adjusted linear modelling technique to account for patient covariates, the magnitude of the relationship was small (r-squared = 0.32) and on the basis of the study, TCB is not seen as a realistic technique to monitor ICP in TBI. Target controlled infusion (TCI) of anaesthetic drugs exploit known pharmacokinetic pharmacodynamic (PKPD) models to achieve set concentrations in the plasma or an effect site. Following a discussion of PKPD model development for the anaesthetic drug propofol, the Validation Study of the Covariates Model (VaSCoM) describes a joint PKPD study of the Covariates Model. Pharmacokinetic validation of plasma concentrations predicted by the model in forty patients undergoing general anaesthesia confirmed a favourable overall bias (3%) and inaccuracy (25%) compared to established PKPD models. The first description of the pharmacodynamic behaviour of the Covariates Model is provided with an estimated rate constant for elimination from the effect site compartment (ke0) of 0.21 to 0.27 min-1.

The definition of awareness used consistently in this thesis is: Explicit recall of intraoperative events during general anaesthesia. Since there is no objective method to detect awareness, the patients must be interviewed after anaesthesia. The form and timing of the interview is crucial. To rely on spontaneous disclosure of awareness episodes is not sufficient. The total number of awareness-victims is considerable although the incidence may seem modest. A number of these patients look upon the awareness experience as the worst experience in their life. Suffering can include pain, mental distress and delayed psychological symptoms. However, the experience of awareness is not uniform and not all patients suffer. A comprehensible definition for dreaming during anaesthesia is: Any recalled experience, excluding awareness, which occurred between induction of anaesthesia and the first moment of consciousness upon emergence. Some findings point in the direction that dreaming during anaesthesia may be related to light or insufficient anaesthesia, but other findings do not. Some patients find dreaming during anaesthesia distressing, but generally the overall impression is that consequences of dreaming during anaesthesia seem to be small and of minor importance to the majority of patients. In this thesis I have found the following:The incidence of awareness is approximately 0.2% when neuromuscular blocking drugs are used and awareness also exists without these drugs, albeit to a lesser extent. These findings represent standard practice in an adult population at normal risk. 50% of awareness cases may have delayed recall of awareness. Using a consecutive inclusion design we found initial awareness suffering comparable to previous studies, but a lower incidence and less pronounced severity of late psychological symptoms. The incidences found among the awareness-victims in our study were; experience of pain 46%, immediate mental distress 65%, any late psychological symptom 33%, and PTSD below 10%. A memory of an intraoperative dream after general anaesthesia is not an early interpretation of delayed awareness, indicating that no routine follow up of dreaming-only patients is indicated. Dreams reported after anesthesia are generally not related to insufficient anesthesia defined as high BIS, and should not be regarded as near awareness.

Volumetric CO2 data from patients in anaesthesia delivery systems are sought after by physicians. The CO2 data obtained with the commonly used sidestream sampling technique are not considered adequate for volumetric CO2 estimation due to distortion and desynchrony with patient flow. The purpose of this thesis was to explore the possibility of using signal enhancing methods to the sidestream data to accurately estimate CO2 flow using a Flow-i anaesthesia delivery system. To evaluate sidestream performance, experimental data was acquired using a mainstream and a sidestream capnograph connected in series to a FRC test lung with known CO2 content, ventilated by a Flow-i anaesthesia machine. The data was then enhanced and analysed using signal processing methods including sigmoid modelling and neural networks. A Feed Forward Neural Network achieved results closest resembling the mainstream capnogram of the evaluated signal processing methods. The mainstream capnogram, considered the benchmark, produced large internal scattering and approximately 25 % offset from actual CO2 flow while using the inherent patient flow data produced by the Flow-i anaesthesia system. When using patient flow data from a Servo-i ventilator, the resulting CO2 flow estimates were drastically improved, producing estimates within 10 % error. This thesis concludes that there are several potential processing methods of the sidestream data to approximate the mainstream signal, however the patient flow of the Flow-i system are a suspected source of error in the CO2 flow estimation.

Introduktion: Anestesi på barn ställer anestesisjuksköterskan inför en komplex situation. Barn är oroliga inför det okända som ska hända. För att minska oron behöver anestesisjuksköterskan reducera deras nivå av oro. Föräldrarna tryggar barnen och tillsammans med dem ska anestesisjuksköterskan delge barnen trygghet. Syfte: Beskriva anestesisjuksköterskors erfarenheter i att bemöta och lindra oro hos barn som ska genomgå anestesi. Metod: En kvalitativ intervjustudie, innehåll analyserades utifrån kvalitativ innehållsanalys. Sammanlagt intervjuades 8 anestesisjuksköterskor med varierande ålder och yrkeserfarenhet. Resultat: Anestesisjuksköterskornas erfarenheter visade att skapa trygghet, ha ett avledande tillvägagångsätt och att utgå ifrån den egna professionen var viktigt för att bemöta och lindra oro. Att samarbeta med föräldrarna, skapa en relation till barnet och ha bra preoperativa förberedelser tycktes skapa trygghet hos både barnet och föräldern. De använde sig av avledande strategier för att flytta den negativa oron till något positivt. Den erfarenhet som anestesisjuksköterskorna hade gav dem en trygg grund i de situationer som inte var fullt så optimala. Då det inte fanns några färdiga mallar i hur de ska bemöta och lindra oro hos barn ansåg de flesta att ”fingertoppskänslan” har stor betydelse. Det viktiga var att hela tiden anpassa sig till barnet och ta den tid som behövdes. Slutsats: Studien belyste anestesisjuksköterskans erfarenhet av att bemöta och lindra oro hos barn. I resultatet framkom att anestesisjuksköterskan upplevde svårighet att bemöta oroliga barn samt att det var en svår situation de ställdes inför. Emellanåt kunde samarbetet med oroliga föräldrar vara svårt, dock kunde samarbetet förbättras om anestesisjuksköterskan och föräldern hade samma målbild. Det ställdes höga krav på kompetens och ödmjukhet samtidigt som anestesisjuksköterskan hade en viktig roll i att skapa trygghet hos barnen.
Introduction To prepare a child for anesthesia put the nurses in a complex situation. The child is worried about the unknow that is going to happen and therefore the nurse needs to reduce the child’s level of anxiety and worry. The parents are supposed to comfort the child and together with the nurse they should help the child to manage the difficult situation. Aim: Describe the anesthesia nurses experience in addressing and reducing anxiety in children who are about to undergo anesthesia. Method:A qualitative interview study of 8 nurses with different age and work experience where included and interviewed. Result: Based on the experience of the nurses it showed that the most important thing where to create comfort, use distractions and the ability to use their previous professional experience. To be able to cooperate with the parents, create a relationship with the child as well as excellent preoperative preparation seemed to create comfort in both the child and the parents. The nurses used distracting strategies to transfer the negative anxiety to something more positive.  Since there are no written guidelines of how to comfort children about to undergo anesthesia the nurses had to use and incorporate their previous experience as an anesthesia nurse. The nurses mentioned that it is crucial to be able to sense and evaluate the situation and then act based on the conclusion. The most important thing was to adapt and conform to the need of the child and not rush the process. Conclusion: The study highlighted the experience of the nurse’s preoperative care. The result showed that the nurses experienced difficulty to care for worried children as well as the complex situation they were presented with. Sometimes worried parents made the situation more complex for the anesthesia nurse since they were not able to cooperate and work together. However, the corporation could be improved if the anesthesia nurse and the parent had the same goal and a joint way of handling the situation. The anesthesia nurse were required to have high professional competence, advanced humility as well the ability to make sure that the child  is comfortable and relaxed.

Ovariohysterectomy and castration are common surgical procedures in small animal practice that can result in clinically significant postoperative pain. One way of controlling postoperative pain is administration of a single analgesic or a combination of different classes of analgesics prior to the onset of noxious stimuli. A constraint to the perioperative use of traditional opioids and non-steroidal anti-inflammatory drugs (NSAIDs) is their undesirable side effects. In this series of experiments, the preventive (pre-emptive) analgesic efficacy of two popular human analgesics, tramadol (an ?atypical? opioid) and parecoxib (a NSAID with selective COX-2 inhibition) was evaluated in dogs. Initially, the efficacy and renal safety of parecoxib, tramadol and a combination of parecoxib, tramadol and pindolol (a -adrenoceptor blocker and 5-HT1A/1B antagonist) were screened in anaesthetised healthy dogs. These analgesics increased the dogs? nociceptive threshold to mechanical stimuli, without causing significant alterations in the dogs? glomerular filtration rate (GFR) estimated by plasma iohexol clearance. Subsequently, the efficacy of tramadol was compared with morphine, in dogs undergoing ovariohysterectomy or castration. The Glasgow composite measure pain scale-short form score (CMPS-SF) and changes in intraoperative electroencephalogram (EEG) responses were used to assess the efficacy of analgesics. Of the three treatment groups (preoperative morphine, 0.5 mg kg-1; preoperative tramadol, 3 mg kg-1; a ?combination? of preoperative low-dose morphine, 0.1 mg kg-1, and postoperative tramadol 3 mg kg-1), dogs given the ?combination? had significantly lower pain scores after ovariohysterectomy. In castrated dogs, preoperative tramadol (3 mg kg-1) and morphine (0.5 mg kg-1) were tested and no significant difference in the CMPS-SF score were observed between them. Changes in EEG variables were not specific between the treatment groups in ovariohysterectomised dogs. Finally, the efficacy of test drugs was evaluated against acute noxious electrical stimulation in anaesthetised dogs, using EEG. Median frequency of the EEG, a reliable indicator of nociception, increased significantly in tramadol and parecoxib groups, compared to morphine, after electrical stimulation. These studies demonstrated that tramadol and parecoxib can produce analgesia in dogs with insignificant side effects. The efficacy of tramadol appears to vary with the type of noxious stimulus. A complete prevention of noxious input by administration of analgesics pre- and post-operatively could have important clinical applications.

Maintenance of anaesthesia using the intravenous agent propofol has increased following development of pharmacokinetic models. An analyser capable of determining propofol concentrations at the point of care may lead to an improved accuracy of drug delivery. Validation work on a novel analyser measuring propofol concentration in near real time demonstrate a high level of precision for samples in the clinical range. Further work in the clinical setting was carried out using the novel propofol analyser to further research its potential use in a diverse patient cohort. Studies were performed in intensive care correlating blood propofol concentrations with depth of sedation, demonstrating a correlation with organ failure. The Marsh model of Target Controlled Anaesthesia was poorer at predicting propofol concentration in patients with significant organ dysfunction than in those without organ failure (correlation coefficient 0.36 vs. 0.73 respectively). Studies in the operating room were performed in which measured propofol concentrations were compared with those predicted using the Marsh model. Results demonstrated significant inaccuracies of the model (bias 32%, precision -8.7 to 72.6%). A method of Marsh model bias correction using a single blood propofol measurement was tested. Results demonstrated insufficient predictability to allow a single point calibration.

The objectives of this thesis are : 1) To provide a review of the literature on the significance, pathogenesis, diagnosis and management of secondary infections in the Intensive Care Unit. 2) To present the findings of a study of the technique of selective parenteral and enteral antisepsis regimen (SPEAR) in the patient population of the Respiratory ICU at Groote Schuur Hospital, aimed at reducing the incidence of secondary infection and, further to evaluate the study in terms of the effect of SPEAR on the incidence of secondary infection and its influence on the mortality due to secondary infection. 3) To present the findings of the effect of SPEAR on patient bacterial colonisation in the ICU, and to evaluate its longterm influence on the microbial flora of the ICU.

This study consisted of two stages. Stage 1 investigated the reproducibility and practicality of two observational sedation assessment scales for use in critically ill children. The two scales were different in design, the first being simple in design requiring a single assessment of the patient. The second was more complex in design requiring assessment of five patient parameters to obtain an overall sedation score. It was established that nursing staff preferred the second, more complex sedation scale mainly because it was perceived to give a more accurate assessment of level of sedation and anxiety rather than merely level of sedation. Stage 2 investigated the pharmacokinetics and pharmacodynamics of midazolam in critically ill children. 52 children, aged between 0 and 18 years were recruited to the study and 303 blood samples taken to analyse midazolam and its metabolites, 1-hydroxymidazolam (1-OH) and 4-hydroxymidazolam (4-OH). A significant correlation was found between midazolam plasma concentration and sedative effect (r=0.598, p=0.01). It was found that a midazolam plasma concentration of 223ng/ml (±31.9) achieved a satisfactory level of sedation. Only a poor correlation was found between dose of midazolam and plasma concentration of midazolam. Similarly only a poor correlation was found between sedative effect and dose of midazolam. Clearance of midazolam was found to be 6.3ml/kg/min (±0.36), which is lower than that reported in healthy children (9.11-13.3ml/kg/min). neonates produced the lowest clearance values (1.63ml/kg/min), compared to children aged 1 to 12 months (8.52ml/kg/min) who achieved the highest clearance values. Clearance was found to decrease after the age of 12 months to values of 5.34ml/kg/min in children aged 7 yeas and above. Patients with renal (n=5) and liver impairment (n=4) were found to have reduced midazolam clearance (1.37 and 0.74ml/kg/min respectively). Disease state was found to affect production of 1-OH. Patients with renal impairment (n=5) produced the lowest 1-OH midazolam plasma ratio (0.059) compared to patients with head injury (0.858).

Background: Neonates who require specialized care and life-saving therapies in neonatal intensive care units (NICUs) and neonatal special care units (NSCUs) can be exposed to separation from their parents and families. Consequently, establishing a parental-neonate bond can be difficult. However, addressing this problem of separation through involving parents and families in neonatal care to improve parent-professional collaboration can result in positive outcomes for neonates and their families. Family-centred care (FCC) has developed over decades and is broadly recommended as an ideal model of care in daily clinical practice in NICUs. However, FCC implementation is challenging at individual, organizational, cultural, and healthcare system levels. In particular, developing countries are challenged by the lack of material resources, infrastructure, and staff shortages. In Thailand, the practical incorporation of FCC into daily clinical practice in neonatal care units is difficult, and it has not been sustainably achieved. Furthermore, there has been minimal research reporting on the development, implementation, and evaluation of FCC in the neonatal critical care context within Thailand. Aim and Objectives: The aim of this PhD study has been to develop, implement and evaluate innovation to facilitate FCC by improving respect, collaboration, and support in a Thai NICU. It had three objectives, each representing a distinct phase in the study: (1) to identify perceptions, current practices and FCC strategies; (2) to develop and implement an innovation to facilitate FCC by improving respect, collaboration, and support in a Thai NICU; and (3) to evaluate the FCC innovation developed in Phase 2. Methods and Results Design: The multistage, mixed-methods study design applied the Participatory Intervention Model (PIM) to guide the innovation’s development, implementation, and evaluation to facilitate FCC by improving respect, collaboration, and support in a Thai NICU. Setting and context: This study was conducted in a tertiary care hospital in southern Thailand (February 2020-January 2021). Ethics approval was obtained from the Research Ethics Committee of Hatyai Hospital and Griffith University. Phase 1: Identification of perceptions, current practices, and FCC strategies Phase 1 was planned to include data collection over 3 months. Due to the COVID-19 pandemic, it was reduced to 2 months during the very early stages of the pandemic (February to March 2020). This phase consisted of two parts, including surveys and interviews with parents and the interdisciplinary professionals. Participants: Participants consisted of two groups: parents of neonates (all gestational ages with no life-threatening or life-limiting diagnosis) who had an expected NICU stay of at least 72 hours and visited the study NICU at least once, and interdisciplinary professionals with a permanent position for at least 1 year in the study unit. Part A: Survey of parents and interdisciplinary professionals Surveys of parents and interdisciplinary professionals were conducted using the validated Perceptions of Family Centred Care – Parent (PFCC-P) and Perceptions of Family Centred Care – Staff (PFCC-S) instruments which were translated into Thai. Sample size: Sample size was based on availability of parents and interdisciplinary professionals over the planned 3-month Phase 1 period. Recruiting parent participants in Phase 1 was prior/during the very early stages of the COVID-19 pandemic and needed to be stopped prior to pre-determined sample size of 100 parents due to visitor restriction (n = 85). Eighty-five parents and 20 interdisciplinary professionals completed the surveys. Data analysis: Demographic characteristics of parents, interdisciplinary professionals, and neonates are reported using descriptive statistics. The subscale scores for parents and interdisciplinary professionals were not normally distributed, so medians were calculated for each of the three sub-scales (respect, collaboration, and support). Parents’ and interdisciplinary professionals’ perceptions of FCC (PFCC-P & PFCC-S) were compared using the Mann-Whitney U test to examine differences in medians in the preimplementation phase because they were unpaired groups. Part B: Semi-structured interviews with parents and interdisciplinary professionals Face-to-face, semi-structured, individual interviews were planned to gain information from extended family members and parents and interdisciplinary professionals; however, given the visitation restrictions, only parents and interdisciplinary professionals were recruited to participate (during the first half of February 2020). Sample size: The sample size was determined when data saturation was identified. Eight interdisciplinary professionals and nine parents participated in face-to-face interviews. Data analysis: Thematic analysis was used to analyse the transcribed Thai language interviews. Results: The survey results across the median of three subscales demonstrated that parents and interdisciplinary professionals’ perceptions on the FCC strategies in current practice were 2-3/4 (Interquartile range [IQR] 1.7-3.8) and 3-4/4 (2.85-3.55), respectively. Considering the median subscale scores, the interdisciplinary professionals had significantly higher subscale scores for respect (median 3.00 (95% CI, 2.91-3.24) vs 2.50 (2.37-2.81)), collaboration (median 3.22 (3.10-3.37) vs 2.33 (1.9-2.62)), and support (median 3.20 (3.03-3.39) vs 2.60 (2.03-2.61)) (all p ≤ 0.001). The interview findings highlighted that the interdisciplinary professionals in this study accepted that the three critical elements of FCC (respect, collaboration, and support) were necessary to be implemented into clinical practice. However, they believed that in reality it was not easy in the Thai NICUs context. This finding identified that the challenge to promote parent-healthcare professional partnerships was associated with the structure and processes of the healthcare delivery system. In addition, the individuality of families' readiness and healthcare providers' perceptions of parents’ involvement as obstacles to providing care were found to be challenges to current practices of FCC. Phase 2: Development and implementation of innovation to facilitate FCC This phase was achieved by two different methods: strategy development working group and implementation of the FCC innovation. Strategy development working group: The development of FCC innovations by the strategy development working group (June to August 2020) was based on Phase 1 findings and the reported integrative literature review. In addition, the FCC innovations were considered within the policies and practices of the NICU in the context of COVID- 19 in Thailand. The development working group members were key and high-level stakeholders in the NICU. Educational activities for the healthcare professional team to incorporate the FCC innovations into their clinical practice in NICU were provided. Implementation of the FCC innovation: The FCC innovations were then implemented over 2 months (September to October 2020), during a period of restrictions on parents and staff arising from COVID-19. Results: The working group identified the gaps in the three key elements (respect, collaboration, and support) to providing FCC in a Thai NICU through the analysis of Phase 1’s results in consort with the findings from the integrative review. A preliminary protocol for the FCC innovations and implementation plan were developed consistent with the challenges associated with COVID-19 in Thailand. FCC practice innovations associated with improving communication were established, including changes and updates to the material within the parent booklet with specific material related to COVID- 19, neonatal updates at bedside or conducted via telephone calls, interdisciplinary family meeting for complex care situations, structured communication checklists, and documentation templates. In addition, although visiting restrictions were limiting, parents were provided with more flexibility as to when they could visit based on individual circumstances. The majority of the healthcare providers in this setting (80%) attended the educational activities to incorporate the FCC innovations into their clinical practice in the NICU. The FCC innovations were incorporated into daily NICU practice by nurses in cooperation with other healthcare providers and ancillary support staff during the pandemic. Phase 3: Evaluation of the FCC innovation Phase 3 (post-implementation) was conducted over 3 months (November 2020-January 2021), and it focused on evaluating the FCC innovations. This phase repeated the collection of data from the validated PFCC-P and PFCC-S surveys of parents and interdisciplinary professionals' perceptions, as per Phase 1, to assess respect, collaboration, and support changes after implementing the FCC innovations in the Thai NICU during the pandemic. Sample size: One hundred parents and 20 interdisciplinary professionals completed the surveys. Data analysis: As per Phase 1 for demographic characteristics. The Mann-Whitney U test was used to analyse parents' perceptions of the items of the PFCC-P pre- and postimplementation given they were two independent groups. Wilcoxon signed-rank test was used to compare the perceptions of the interdisciplinary professionals pre- and postimplementation using the PFCC-S given they were matched samples. Results: The participants consisted of 83 pairs of parents (i.e., mother and/or father of neonate participated) (35 pre; 48 post), which represented 102 neonates (50 pre; 52 post). There were 185 parents; 85 pre-implementation and 100 post-implementation. For the NICU health care team, 20 participated. The median scores of parents' perceptions post-implementation significantly improved for respect (2.50 to 3.50; 95%CI, 3.02-3.53), collaboration (2.33 to 3.33; 2.90- 3.40), support (2.60 to 3.60; 2.84-3.62), and the overall score (2.50 to 3.43) (p < 0.001, 95%CI 2.93-3.51). There was an absolute difference of at least 0.3 in the pre- and postimplementation scores for three subscales and overall score, where 0.3 corresponds to 10% of the rating scale. Comparatively, interdisciplinary professionals' perception of FCC did not significantly change pre- and post-implementation for respect ([median] 3.00 to 2.92; 95%CI, 2.87-3.16), collaboration (3.22 to 3.33; 3.16-3.47), support (3.20 to 3.20; 2.96-3.28) and overall (3.15 to 3.20; 95%CI, 3.10-3.25). Conclusions:Results from this study indicate that incorporating FCC innovations in the NICU appeared to be successful, despite the challenges of COVID-19. The key finding was that the innovations incorporated in the NICU were primarily based on communication strategies, a simple means to support, collaborate with, and respect parents that required low investment within the complex situation arising from COVID-19. These innovations were essential to engage collaborative working between parents and healthcare providers to promote parents as partners in a neonatal critical care team. To successfully implement FCC innovations in different settings, further innovations associated with communication methods need to target the specifics of individuals involved, healthcare settings, and available resources.
Thesis (PhD Doctorate)
Doctor of Philosophy (PhD)
School of Nursing & Midwifery
Griffith Health
Full Text

Palliative health care is offered to any patient experiencing a life limiting or life changing illness. The palliative approach includes goals of care, expert symptom management, and advance care planning in order to reduce patient suffering. Complex care can be provided by palliative care specialists while primary palliative care can be given by educated staff nurses. However, according to the literature, intensive care unit (ICU) nurses have demonstrated a lack of knowledge in the provision of primary care as well as experiencing moral distress from that lack of knowledge. In this doctor of nursing practice staff education project, the problem of ICU nurses' lack of knowledge was addressed. Framed within Rosswurm and Larrabee's model for evidence-based practice, the purpose of this project was to develop an evidence-based staff education plan. The outcomes included a literature review matrix, an educational curriculum plan, and a pretest and posttest of questions based on the evidence in the curriculum plan. A physician and a master's prepared social worker, both certified in palliative care, and a hospital nurse educator served as content experts. They evaluated the curriculum plan using a dichotomous 6-item format and concluded that the items met the intent of the objectives. They also conducted content validation on each of the pretest/posttest items using a Likert-type scale ranging from 1 (not relevant) to 4 (very relevant). The content validation index was 0.82 indicating that test items were relevant to the educational curriculum objectives. Primary palliative care by educated ICU nurses can result in positive social change by facilitating empowerment of patients and their families in personal goal-directed care and reduction of suffering.

Background: Careful airway management, including tracheal intubation, is important when performing anaesthesia in order to achieve safe tracheal intubation. Aim: To study airway management in anaesthesia care from both the professional and patient perspectives. Methods: 11 RNAs performed three airway tests in 87 patients, monitored in a study-specific questionnaire. The tests usefulness for predicting an easy intubation was analysed (Study I). 68 of 74 anaesthesia departments in Sweden answered a self-reported questionnaire about the presence of airway guidelines (Study II). 20 anaesthesiologists were interviewed; a phenomenographic analysis was performed to describe how anaesthesiologists' understand algorithms for management of the difficult airway (Study III). 13 patients were interviewed; content analysis was performed to describe patients' experiences of being awake fiberoptic intubated (Study IV). Results: The Mallampati classification is a good screening test for predicting easy intubation and intubation can be safely performed by RNAs (Study I). The presence of airway guidelines in Swedish anaesthesia departments is poorly implemented (Study II). Algorithms can be understood as law-like rules, a succinct plan to follow in difficult airway situations, an action plan kept in the back of one's mind while creating flexible and versatile personal algorithms, or as consensus guidelines based on expert opinion in order to be followed in clinical practice (Study III). One theme emerged describing experiences of being awake intubated; feelings of being in a vulnerable situation but cared for in safe hands, described in five categories: a need for tailored information, distress and fear of the intubation, acceptance and trust of the staff's competence, professional caring and support, and no hesitation about new awake intubation (Study IV). Conclusions: The Mallampati classification is a good screening test for predicting easy intubation, when the airway assessment is performed in a structured manner by RNAs. The presence of airway guidelines in Swedish anaesthesia departments was poorly implemented and should receive higher priority. Algorithms need to be simple and easy to follow and based on the best available scientific evidence. Tailored information about what to expect, ensuring eye contact, and giving breathing instructions during the procedure may reduce patients' feeling distress.

Ventilator-associated pneumonia (VAP), catheter-related blood stream infection (CRBSI) and pressure ulcers (PU) are well recognized complications in intensive care units (ICUs). Many of these are preventable but can also complicate patient recovery, prolong length of stay, increase costs, morbidity and mortality. In Malaysia, the majority of studies investigating VAP and CRBSI in Malaysia have focussed on identifying risk factors, diagnostic criteria and treatment of ICU-related complications. Further, in spite of the burden of PU there are limited studies undertaken in Malaysia and few of these have been nurse-led. Importantly, to date there has been limited investigation of the efficacy and effectiveness of quality improvement initiatives and the contextual issues impacting on clinical practice improvement in Malaysia.In spite of the increasing emphasis on quality assurance in Malaysian ICUs there has been a limited focus on nurse-specific interventions and the majority of projects have been initiated by physicians. This study has evaluated the utility of a nurse-led action research project to drive clinical practice improvement in the ICU and is significant in demonstrating the capacity of nurses to critique and control their practice. The project conducted for this thesis was called the Improving health outcomes by preveNting intensiVe care related infEction in Malaysia intenSive care uniT - INVEST study. The INVEST Study as reported in this thesis has been undertaken using an action research approach to improve the uptake of evidence-based strategies to prevent infection in the ICU in the Malaysian cultural context.The aims of this thesis were to identify best practices, evaluate the current nursing practice in prevention of VAP, CRBSI and PU in ICU patients in a single Malaysian ICU, and evaluate the impact of the evidence-based interventions to improve patient outcomes. The specific and research objectives of this study were to:1. Identify best practice interventions for preventing VAP, CRBSI and PU in the ICU. 2. Document the current rates of VAP, CRBSI and PU in an ICU in Malaysia. 3. Implement an action research intervention to collaboratively develop and implement strategies for improvement 4. Assess the impact of the intervention on clinical outcomes, staff dynamics, work place culture and sustainability of practice change An action research approach was used in this study to involve and empower nurses and drive practice change. A literature review identified that many action research studies conducted in the ICU were mainly most focused on process measures and not outcomes. In this study the data were collected in three phases following the action research cycles which comprised of a period of planning, acting, observation, reflecting and re-planningIn Phase I of the thesis current best practice interventions for the prevention of VAP, CRBSI and PU in ICU are described. A literature search was conducted to identify evidence-based practices (EBP) that were recommended by bodies to improve the prevention of VAP, CRBSI and PU. A core set of nursing activities was identified in preventing the complications of VAP, CRBSI and PU. These were hand washing, hygiene care, positioning of patient, elevation of the head of bed and providing adequate nutrition.Pre- intervention data collection consisted of an environmental scan, including interview with the key stakeholders, patient profiling and a nurse survey. Twenty-one cases of ICU complications were identified in 18 of the 91 patients (19.8%) admitted in December 2009. Of the patients, three developed two complications - PU and VAP (two patients) or CRBSI (one patient). The findings indicated that this ICU had a high case load due to the high ICU bed demand. Patients needing ICU care were being nursed in general wards due to the unavailability of ICU beds.Nurses reported a good knowledge of prevention strategies with a mean score of 124.84 ±SD14.66 and reported a high level of positive regard for their professional practice environment based on the results of Revised Professional Practice Environment (RPPE). Three components had mean scores of ≥3 and five <3 within the eight components. Three components of RPPE subscales with highest mean scores were Internal Work Motivation (M 3.24; SD 0.3), Relationship With Physician (M 3.22; SD 0.53) and Cultural Sensitivity (M 3.04; SD 0.24). The two lowest mean scores were for Handling Disagreement and Teamwork with 2.77 (SD 0.16) and 2.45 (SD 0.47), respectively. Nurses also showed positive attitudes toward the sustainability of the change process. The Sustainability Indices ranged from 13.4 to 100 with a mean of 75.21 (SD 21.71).In Phase 2 the intervention was conducted over six months from February to July 2010. The Center of Disease Control and Prevention (CDC) criteria for diagnosis of VAP and CRBSI, and the Waterlow Pressure Ulcer Risk Assessment Scale were promoted in the unit. Nurses were exposed and encouraged to implement evidence-based nursing interventions as identified in care criteria. All nurses were invited to the unit nursing education to increase their knowledge and awareness about evidence-based practice in prevention of the ICU complications. Nurses were encouraged to gain control of their practice. Evidence-based practice articles were also provided to increase their knowledge level and posters were distributed and placed in the unit to increase nurses awareness of the quality improvement initiativesFocus group discussions were conducted in Phase 2 and found that nurses in the unit were unaware of the importance of standardized assessment in their daily practice. They had a lack of understanding regarding the importance of standardised risk assessments. Despite the reluctance of many nurses to embrace the EBP, due to a perception of their workload, the focus groups also revealed nurses were optimistic that change will get easier and could be eventually achieved. Participants were positive about the change that could take place in the future. The hierarchical relationships with medical doctors were also identified as a factor limiting nurses from adopting the guidelines.Phase 3 of the project, the post-intervention phase was conducted from March to May 2011. The data collection process was repeated as Phase 1 and Phase 2. There were 11 cases of ICU complications identified during the post-intervention phase in 10 (8.7%) of the 115 patients admitted during March 2011. One patient developed both VAP and PU, while four developed VAP and another five PU. In the post-intervention group, no cases of CRBSI were detected. The total mean score of nurses’ knowledge was 121.45±SD16.85. An independent-samples t-test was conducted to compare nurses’ knowledge pre and post intervention, and found no significant differences, t (150) =1.32, P 0.189. The Sustainability Indices ranged from 41.3 to 100 percent with a mean of 76.81±SD21.45.Approximately 84% of the nurses in pre-intervention and 70% in post-intervention scored >55%. The nurses reported a positive regard for their practice environment in the pre- and post-intervention groups. The mean scores for each component were comparable for both the pre- and post-intervention groups except for Internal Work Motivation, Control Over Practice and Staff Relationship With Physician. The highest mean scores within the eight components for the post-intervention group were for Internal Work Motivation (M 3.13; SD 0.27), Relationship With Physician (M 3.04; SD 0.33) and Cultural Sensitivity (M 3.01; SD 0.23). The three lowest were for Handling Disagreement and Conflict (2.80; SD 0.20), Control Over Practice (2.71; SD 0.34) and Teamwork (2.48; SD 0.31).There was a reduction in overall complications from 19.8% to 8.7%. Few nurses in the focus group were optimistic that at least some changes had taken place, and positively improving their knowledge on assessment of patients and some of their common practices in the ICU. The challenge, which they were presently facing was the implementation of hospital information system because most of them were not knowledgeable in information technology.The main outcome of this study was that there was a reduction in number of patients with PU from 16 to 6 in pre and post intervention groups. This reduction of PU was statistically significant (χ[superscript]2=8.14, df=1, p=0.04).In conclusion whether there was a real improvement in patient care provided due to the interventions given was not able to be determined due to methodological considerations and inability to control for confounders. These data underscore the importance of considering cultural factors, both organisational and societal in quality improvement initiatives and empowering nurses for practice change. A risk management system which acknowledges competing demands in dynamic, real world environments is important to consider in future quality improvement studies. The series of studies presented in this thesis have contributed to understanding of factors influencing implementation and sustainability of quality improvement initiatives in a Malaysia ICU. Information acquired from the thesis will be useful information for further improvement targeting education, services, research, policy and future quality improvement project plans in Malaysia.

Computer and Information Science
M.S.
Many emerging and existing medical applications can benefit from having continuous access to the patients vitals. This paper presents the results of a set of experiments conducted in a medical setting to determine the feasibility of using wireless communication in both home care and hospital intensive care environments. The study is also done with the intention of developing a new wireless protocol for use in medical settings. This protocol will later be incorporated into different medical devices operating inside a patient room with a view that significant performance improvements should be observed.
Temple University--Theses

Kangaroo care (KC) is a cost-efficient method to increase infant-parent bonding and neonatal health outcomes worldwide. Despite evidence supporting KC in critically ill infants, nursing perceptions regarding patient safety and interrupted work flow continued to impede practice in the local high-tech neonatal intensive care unit (NICU). Their current policy failed to address the 2-person transfer method recommended for safe practice. In addition, both staff and parents lacked training and education regarding the benefits and feasibility of KC. This doctoral project aimed to decrease practice barriers and promote earlier and more frequent KC by developing and integrating an evidence-based clinical pathway within a multifaceted champion-based simulated educational training program for NICU staff and parents. Published outcomes and generated organizational data for program synthesis connected the gap in practice. Kolcaba's comfort theory served as the guiding framework to ensure a partnership in care. This quasi-experimental quantitative study used the generalized liner model for data analysis. Study findings indicated that KC occurred 2.4 more times after the intervention compared to before (p = 0.001). Descriptive data revealed that KC episodes for intubated patients nearly doubled after implementation (11.1% from 6.2%). Post-survey scores for nursing knowledge and comfort level also improved after the intervention. Although earlier KC practice was non-conclusive (p = 0.082), future trials should control groups for day of life since admission. Disseminating the KC pathway can have a positive social change on family-centered care by increasing NICU nurses' knowledge, comfort, and adoption of this evidence-based practice as an expected routine standard of care.

Nursing Care for Terminal Patients in Intensive Care Units by Pervell Dunbar Project Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Nursing Practice Walden University August 2015 Although the goal of the ICU has always been to save lives, ICU now additionally provides end-of life (EOL) care. The objective of this project was to provide ICU nurses with a comprehensive awareness of physical, emotional, and spiritual EOL care issues of patients and their families in order to be better equipped to handle EOL care. The framework used was Jean Watson's Caring model (10 Caritas). A literature review revealed a poster previously used by a major health organization as a conversation starter to facilitate decision-making among ICU nurses, EOL patients, and their families related to EOL issues. The purpose of this quality improvement initiative was to introduce and implement an educational EOL tool that would engage patients and family members in meaningful and useful conversations with ICU nurses. Twenty seven ICU nurses were selected by the unit's director to attend a PowerPoint presentation on the use of the EOL educational poster. Four ICU nurses were chosen by the director to be champions for this project. After the presentation, there was a period for questions and answers, and the ICU nurses were requested to give feedback on the presentation. The result from the feedback revealed that EOL care is outside previous practice and may require extra education and support. These comments substantiated similar conclusions from other researchers as described in this paper. With an increase in EOL training for ICU nurses and the implementation of EOL teaching tools like the poster used in this study, ICU nurses may be better able to have conversations with EOL patients and families, thus improving patient care.

Jordan is a developing country which is taking measures to make the situation in the society more equal between males and females. Former research has showed that it, worldwide, sometimes is great differences between the genders in the health care. This thesis illuminates how it is to work in an intensive care unit and if there are differences between male and female intensive care nurses influencing on the provided care. Twenty intensive care nurses from four intensive care units at Jordan University Hospital were included in the study. Both field notes and an observation schedule were used to gather data. The field notes captured the overall experience of working in the intensive care area and were analyzed by thematic content analyze. The observation schedule concentrated on how long time was spent and which activities were performed bedside. Data from the schedule was compared between the units and between male and female intensive care nurses caring for male and female intensive care patients. The field notes showed that even if the units were busy and crowed the silence and calmness were present. Cooperation was essential and trust and knowledge were spread. Often a warm and comfortable feeling surrounded the personnel and they seemed to like it at work. The observation schedule showed that in the medical and surgical intensive care units the intensive care nurses spent more time and performed a greater number of bedside activities compared to the pediatric and main intensive care units. Female intensive care nurses who cared for male intensive care patients performed less bedside activities and when they cared for a female intensive care patient they spent less time bedside compared to the other groups. In all the units and all the groups the most frequent performed activity was to have a look at the equipment, followed by have a look at the patient. The results can however be questioned since it is a small study accomplished by a single researcher, in some few intensive care units at one hospital.
Program: Specialistsjuksköterskeutbildning med inriktning mot intensivvård

Background: Postoperative critical care support is required for emergency and elective cases having either major surgery, with poor physiological states or significant comorbidities, and for support following unexpected surgical or anaesthetic complications. Research suggests that as many as 48% of all critical care unit (CCU) admissions occur postoperatively, yet limited literature is available regarding the support role that onsite critical care availability provides for surgery. Research into this area is therefore necessary to understand the impact of accessible critical care support, especially in hospitals at regional and district level. Objectives: The objective of this research is to contribute to the literature on perioperative critical care by presenting data quantifying and describing the patients requiring postoperative critical care at New Somerset Hospital (NSH) - a regional hospital in Cape Town, in the Western Province of South Africa. Further to this, the research aims to identify cases that would not have proceeded here if the option of on-site postoperative critical care did not exist. Methods: Data was collected using a prospective survey spanning a six-month period from June 2015 to November 2015. The data represented two sets of patients: 1) every case done, documenting whether they would have proceeded at NSH without the presence of a critical care unit; 2) each admission to a critical care service directly from theatre, describing their indications for admission and their postoperative critical care pathway, interventions and outcomes. Results: A total of 3247 complete cases were included in the analysis. Of the total sample of cases assessed, 66 (2%) were supported by critical care at NSH, of which roughly half (31 cases) would not have proceeded at NSH without availability of a critical care bed. Of these patients, 7 did not have a bed reserved preoperatively, and were not admitted, highlighting an important subgroup of patients: those not admitted to a CCU, but yet received surgery at NSH solely due to the potential of postoperative critical care support there. New admissions amounted to 48 (1.5%) of all cases of which 43 were emergencies, and 14 were unplanned. 45% of admissions required monitoring or epidural care only, for which High Care would have been sufficient, while 55% received cardiorespiratory support. Conclusion: These results confirm that at NSH, an on-site CCU allows for cases to proceed that would otherwise have been transferred elsewhere. Of note, obstetrics accounted for 3 of the unplanned admissions, confirming that a level 2 obstetric service requires critical care support despite treating otherwise low risk patients. This data indicates that critical care plays a beneficial role in supporting a regional theatre service. 6 Further research is required in this field to determine whether these results can be generalised to other regional hospitals. This survey should help as baseline data, especially for studies to better assess quality and outcomes against national and international metrics.

This thesis reviews the requirements for intensive care audit data and describes the development of ICARUS (Intensive Care Audit and Resource Utilisation System), a system to collect and analyse intensive care audit information. By the end of 1998 ICARUS contained information on over 45,000 intensive care admissions. A study was performed to determine the accuracy of the data collection and entry in ICARUS. The data in ICARUS was used to investigate some limitations of the APACHE II severity scoring system. The studies examined the effect of changes in physiological values and post-intensive care deaths, and the effect of casemix adjustment on mortality predicted by APACHE II. A hypothesis is presented that excess intensive care mortality in the United Kingdom may be concealed by intensive care mortality prediction models. A critical analysis of ICARUS data was undertaken to identify patient groups most likely to benefit from intensive care. This analysis revealed a high mortality in critically ill patients admitted from the wards to the intensive care unit. To help identify critically ill ward patients, the physiological values and procedures in the 24 hours before intensive care admission from the ward were recorded: examination of the results suggested that management of these patients could be improved. This led to the setting up of a patient at risk team (PART). Two studies report the effect of the PART on patients on the wards and on the patients admitted from the wards to the intensive care unit. Additional care for surgical patients on the wards is suggested as a way of improving the management of high-risk postoperative patients. The thesis concludes by discussing the benefits of the ICARUS system and speculating on the direction that should be taken for intensive care audit in the future.

Abstract Background: Pilot studies applying point-of-care ultrasound (POCUS) in preeclampsia indicate the presence of pulmonary interstitial edema, cerebral edema, and cardiac dysfunction. Laboratory markers of oncotic pressure (albumin) and cardiac dysfunction (brain natriuretic peptide [BNP]) may be abnormal, but the clinical application remains unclear. We investigated the prevalence of pulmonary interstitial syndrome (PIS), cardiac dysfunction, and increased optic nerve sheath diameter (ONSD) in late onset preeclampsia with severe features. The primary aim was to examine the association between PIS or ONSD and maternal serum albumin level. The secondary aims were to explore the association between cardiac dysfunction and PIS, ONSD, BNP, and serum albumin level, and between POCUS-derived parameters and a suspicious or pathological cardiotocograph (CTG). Methods: Ninety-five women were enrolled in this prospective observational cohort study. A POCUS examination of lungs, heart and ONSD was performed. PIS was defined as a bilateral B-line pattern on lung US, and diastolic dysfunction according to an algorithm of the American Society of Echocardiography. ONSD > 5.8 mm was interpreted as compatible with raised intracranial pressure (> 20 mmHg). Serum BNP and albumin levels were also measured. Results: PIS, diastolic-, systolic dysfunction, and raised left ventricular end-diastolic pressure (LVEDP) were present in 23 (24%,) 31 (33%), 9 (10%), and 20 (25%) women respectively. ONSD was increased in 27 (28%) women. Concerning the primary outcome, there was no association between albumin level and PIS (p = 0.4) or ONSD (p=0.63). With respect to secondary outcomes, there was no association between albumin level and systolic dysfunction (p = 0.21) or raised LVEDP (p = 0.44). PIS was associated with diastolic dysfunction (p = 0.02), and raised LVEDP (p = 0.009, negative predictive value 85%). BNP level was associated with systolic (p < 0.001)- and diastolic dysfunction (p = 0.003) and LVEDP (p = 0.007). No association was found between POCUS abnormalities and a suspicious/pathological CTG (p = 0.07). Conclusion: PIS, diastolic dysfunction and increased ONSD were common in preeclampsia with severe features. Cardiac ultrasound abnormalities may be more useful than albumin levels in predicting PIS. The absence of PIS may exclude raised LVEDP. The further clinical relevance of PIS and raised ONSD remains to be established. BNP level was associated with cardiac ultrasound abnormalities. Although this study was not designed to directly influence clinical management, the findings suggest that POCUS may serve as a useful adjunct to clinical examination for the obstetric anesthesiologist managing these complex patients.

Abstract Systematic analyses of infections in critical illness are sparse and mostly restricted to specific infection categories. Thus, a prospective study was carried out in a medical-surgical ICU during 14 months on patients whose ICU stay was longer than 48 h. The prospectively gathered data included detailed patient history, infection survey, severity of illness scores (APACHE II, SOFA), resource use, short-term and long-term outcome and quality of life following hospital discharge. Altogether 335 patients were included, of whom 251 (74.9%) had an infection on admission; 59.3% had a community-acquired infection (CAI) and 40.7% a hospital-acquired infection (HAI), while 84 (25.1%) did not have any infection (NI). APACHE II scores and ICU or hospital mortality rates did not differ between the groups. The median hospital stay was longer in the HAI than in the CAI or NI groups. Eighty (23.9%) of the 335 patients developed an ICU-acquired infection (48 per 1000 patient days): ventilator-associated pneumonia (VAP) in 33.8% of the cases, central catheter-related (CRI) or primary bloodstream infections in 6.3% and urinary tract infections in 1.3%, while the corresponding device-related incidences per 1000 days were 18.8, 2.2 and 0.5, respectively. ICU-acquired infection was an independent risk factor for hospital mortality. It doubled the risk for hospital mortality in patients with an infection on admission and caused a threefold the risk in patients without an infection on admission and an almost fourfold increase in the use of nursing resources. Of the 272 hospital survivors, 83 (30.5%) died after discharge during the median follow-up of 17 weeks. Infection status on admission or during the ICU stay did not affect long-term mortality. ICU-acquired infection did not have an impact on patients' quality of life. The current general level of health compared to the status before ICU admission did not differ between the groups, either. Only 36% of those employed resumed their previous jobs. Three-fourths of patients had an infection on admission, while nearly one fourth acquired an ICU infection. The high VAP rate suggests a need for re-evaluation of preventive measures, whereas the low CRI indicates more successful prevention. ICU-acquired infection was a significant risk factor for hospital mortality, but did not affect patients' long-term survival or quality of life.

The vital signs monitoring data of an infant receiving intensive care are a rich source of information about its health condition. One major concern about the state of health of such patients is the onset of neonatal sepsis, a life-threatening bloodstream infection. As early signs are subtle and current diagnosis procedures involve slow laboratory testing, sepsis detection based on the monitored physiological dynamics is a clinically significant task. This challenging problem can be thoroughly modelled as real-time inference within a machine learning framework. In this thesis, we develop probabilistic dynamical models centred around the goal of providing useful predictions about the onset of neonatal sepsis. This research is characterised by the careful incorporation of domain knowledge for the purpose of extracting the infant’s true physiology from the monitoring data. We make two main contributions. The first one is the formulation of sepsis detection as learning and inference in an Auto-Regressive Hidden Markov Model (AR-HMM). The model investigates the extent to which physiological events observed in the patient’s monitoring traces could be used for the early detection of neonatal sepsis. In addition, the proposed approach involves exact marginalisation over missing data at inference time. When applying the ARHMM on a real-world dataset, we found that it can produce effective predictions about the onset of sepsis. Second, both sepsis and clinical event detection are formulated as learning and inference in a Hierarchical Switching Linear Dynamical System (HSLDS). The HSLDS models dynamical systems where complex interactions between modes of operation can be represented as a twolevel hidden discrete hierarchical structure. For neonatal condition monitoring, the lower layer models clinical events and is controlled by upper layer variables with semantics sepsis/nonsepsis. The model parameterisation and estimation procedures are adapted to the specifics of physiological monitoring data. We demonstrate that the performance of the HSLDS for the detection of sepsis is not statistically different from the AR-HMM, despite the fact that the latter model is given “ground truth” annotations of the patient’s physiology.

Purpose: The purpose of this project was to conduct an educational workshop among ICU nurses working in rural areas, to reduce risk of compassion fatigue. Background: Compassion fatigue is a job-related hazard unique to professionals in caring professions such as nursing. Rural ICU nurses represent a population that may encounter unique triggers for the risk of compassion fatigue due to professional isolation, less resources and more risk of knowing the patient as a community member. A review of literature reveals limited research related to compassion fatigue development in rural ICU nurses. Method: Two educational workshops were conducted among rural ICU nurses (N=3). Workshop content included discussion about symptoms, triggers, and outcomes of compassion fatigue, as well as positive coping strategies. Participants journaled physical and emotional responses to situations such as ethical or moral dilemmas, boundary issues, and aspects of self-care. Each workshop included time to discuss the educational content and participant experiences; the resulting narratives were analyzed for commonalities. Findings: Universally, burnout was viewed as inherent to the profession. All participants recounted past traumatic patient encounters that preoccupies their thinking when in similar situations. Symptoms identified as compassion fatigue included chronic, constant, generalized pain, symptoms of depression, isolation, withdrawal and lack of interest in enjoyable activities. Triggers were prolonged patient hospitalizations due to lack of resources, lack of supporting services, lack of leadership support, unexpected patient loss, witnessing patient trauma and grief, caring for patients who did not seem to really need ICU, social situations surrounding patients 10 and floating to different departments. Increased alcohol intake, sarcasm, and venting were the most reported mitigation strategies among participants. Implications: Compassion fatigue negatively impacts the lives of rural ICU nurses on many dimensions, although it is perceived as inherent to the profession. These participants desired support from nursing leadership and a supportive work environment. Participants expressed they continued to feel compassion, despite experiencing the phenomenon described as compassion fatigue. A less stigmatizing term might better capture the phenomenon now labeled as compassion fatigue.

The rapid growth of medical sciences and technologies created the need to manage data generated by sophisticated medical equipment (e.g. lab results, vital signs, etc.). This class of equipment, especially in the modern Intensive Care Unit (ICU), emits large quantities of latient data which medical staff usually records on log sheets.
This thesis presents a database design that allows abstract definition of data types, and offers a unified view of data during the development phase, distinct levels of data management and a higher degree of system flexibility. This database model is an implementation of a database for a Patient Data Management System (PDMS) developed for use in the ICU of the Montreal Children's Hospital. The PDMS has a variety of application modules that handle and process various types of data according to functionality requirements.

The observed physiological dynamics of an infant receiving intensive care contain a great deal of information about factors which cannot be examined directly, including the state of health of the infant and the operation of the monitoring equipment. This type of data tends to contain both common, recognisable patterns (e.g. as caused by certain clinical operations or artifacts) and some which are rare and harder to interpret. The problem of identifying the presence of these patterns using prior knowledge is clinically significant, and one which is naturally described in terms of statistical machine learning. In this thesis I develop probabilistic dynamical models which are capable of making useful inferences from neonatal intensive care unit monitoring data. The Factorial Switching Kalman Filter (FSKF) in particular is adopted as a suitable framework for monitoring the condition of an infant. The main contributions are as follows: (1) the application of the FSKF for inferring common factors in physiological monitoring data, which includes finding parameterisations of linear dynamical models to represent common physiological and artifactual conditions, and adapting parameter estimation and inference techniques for the purpose; (2) the formulation of a model for novel physiological dynamics, used to infer the times in which something is happening which is not described by any of the known patterns. EM updates are derived for the latter model in order to estimate parameters. Experimental results are given which show the developed methods to be effective on genuine monitoring data.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO
COORDENAÇÃO DE APERFEIÇOAMENTO DO PESSOAL DE ENSINO SUPERIOR
CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO
PROGRAMA DE SUPORTE À PÓS-GRADUAÇÃO DE INSTS. DE ENSINO
A Unidade de Terapia Intensiva (UTI) é um departamento importante dentro do Hospital visto que lida majoritariamente com casos de alta complexidade e gera elevados custos administrativos, o que requer um controle adequado de seus processos. Inconformidades tais como erros em atividades de tratamento e falta de comunicação entre os funcionários são comumente responsáveis pelo baixo desempenho de UTIs e devem ser ajustados para reduzir possíveis danos ao tratamento do paciente. Para avaliar a eficiência de uma UTI, a literatura propõe que sejam estabelecidas métricas que considerem quatro perspectivas - médica ou clínica, econômica, social e institucional – que oferecem uma visão abrangente das atividades (administrativas ou de tratamento) dentro da unidade e seus impactos no pós-tratamento. Entretanto, a avaliação de desempenho em uma UTI não é uma tarefa simples, pois há diversas variáveis a serem consideradas e que podem ser potenciais causas de um mau desempenho. Além disso, não há uma métrica ou indicador padrão-ouro que consegue reter de forma adequadas as informações, sendo que diversas perspectivas devem ser consideradas. Os indicadores mais comuns são A Taxa de Mortalidade Padronizada (Standardized Mortality Ratio, SMR) e o Taxa de Uso de Rescursos Padronizada (Standardized Resource Use, SRU), que contabilizam desfechos de mortalidade (clínicos) e de uso de recursos (econômicos), junto de metodologias propostas para viabilizar a comparação entre diferentes UTIs, identificar de grupos de desempenho e analisar os riscos de mortalidade dos pacientes dentro da unidade, tais como os conceitos de Rankability e Perfis de Risco (Risk Profiles). Além disso, é necessário definir corretamente os desfechos a serem contabilizados em indicadores. Nesse contexto, recomenda-se a combinação de diferentes indicadores e metodologias de forma a complementar e elevar a confiabilidade da análise de desempenho e benchmarking. Com isso, este estudo tem como objetivo analisar um conjunto de UTIs em termos de desempenho quanto à mortalidade e uso de recursos, associando-os com as características das unidades e seus fatores institucionais, para identificar possíveis correlações. A análise foi feita em uma amostra composta por 12.100 pacientes que foram hospitalizados em 116 UTIs, considerando um desfecho em até 60 dias de interação. Este estudo teve como contribuição a combinação de diferentes técnicas e indicadores, e uma discussão a respeito da variabilidade do SMR em comparação à metodologia tradicional. Para este propósito, combinou-se as técnicas da Matriz de Eficiência, Rankability – índice de confiabilidade de um indicador de desfecho, e Perfis de Risco, de forma a obter e avaliar o desempenho de grupos de UTIs. Como resultados, verificou-se que UTIs cuja administração é de domínio Público e que destinam a maioria dos seus leitos ao Sistema Único de Saúde (SUS) brasileiro tiveram mortalidade significativamente alta em relação àquelas de dominínio privado (p-valor menor que 0.05). Além disso, realizou-se um agrupamento das UTIs utilizando quatro diferentes técnicas de clusterização de forma a garantir a máxima confiabilidade do indicador para comparação (Rankability), o que resultou na presença de clusters extremos contendo uma UTI cada, sendo elas a de maior e a de menor SMR, apesar de ambas apresentarem o mesmo conjunto de severidades. Para cada grupo, estimou-se o seu perfil de risco, e verificou-se que pacientes com menor gravidade apresentaram maior variabilidade nos riscos de morte, sendo estes maiores nos grupos com alto SMR e menores em grupos de menor mortalidade, sendo que a dispersão tendeu a ser menor quanto menor for o risco, o que poderia influenciar diretamente no cálculo do SMR. Com isso, por meio de equações matemáticas e simulação por meio de reamostragem, verificou-se que o SMR possui uma limitação em sua escala, que depende diretamente do espectro de gravidade dos pacientes em cada UTI ou grupo de desempenho analisado. O S
Intensive Care Unit (ICU) is an important department within a hospital since it deals mostly with complex cases and it generates the highest amount of costs, thus requiring adequate control on its care treatments. Nonconformities such as poor communication and treatment errors are commonly responsible for a bad performance in ICUs. However, evaluating the performance of an ICU is not an easy task and there are no gold-standard indicators. The most common metrics are the Standardized Mortality Ratio (SMR) and the Standardized Resource Use (SRU), which measure mortality and resource utilization, respectively. Hence, this study aims to analyze different ICUs in terms of mortality, resource use, and institutional factors, combining the methods Efficiency Chart, Rankability and Risk Profile. The analysis was performed considering a total of 12,100 patients in 116 ICUs provided by a clinical trial study. As results, it was verified that most ICUs were from hospitals with public administration (47.41 per cent), which had significantly high lethality rate compared to private hospitals. Four different clustering approaches were tested, which identified similar case-mixes between the best and lower performance groups of ICUs, and a high variability in expected risks for low severity patients. Using a resampling approach, it was evidenced that the mortality indicator varies strongly on low-risk groups of patients, while high-risk patients had a smaller range of SMR values, which may lead to biased conclusions when comparing ICUs with similar mortality and different case-mixes.

Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2011.
Cataloged from PDF version of thesis.
Includes bibliographical references (p. 71-74).
Extensive bedside monitoring in hospital Intensive Care Units (ICU) has resulted in a deluge of information on patient physiology. Consequently, clinical decision makers have to reason with data that is simultaneously large and high-dimensional. Mechanisms to compress these datasets while retaining their salient features are in great need. Previous work in this area has focused exclusively on supervised models to predict specific hazardous outcomes like mortality. These models, while effective, are highly specific and do not generalize easily to other outcomes. This research describes the use of non-parametric unsupervised learning to discover abstract patient states that summarize a patient's physiology. The resulting model focuses on grouping physiologically similar patients instead of predicting particular outcomes. This type of cluster analysis has traditionally been done in small, low-dimensional, error-free datasets. Since our real-world clinical dataset affords none of these luxuries, we describe the engineering required to perform the analysis on a large, high-dimensional, sparse, noisy and mixed dataset. The discovered groups showed cohesiveness, isolation and correspondence to natural groupings. These groups were also tested for enrichment towards survival, Glasgow Coma Scale values and critical heart rate events. In each case, we found groups which were enriched and depleted towards those outcomes.
by Kanak Bikram Kshetri.
M.Eng.

Thesis: M. Eng. in Computer Science and Molecular Biology, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2015.
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 73-74).
We investigated the relationship between initial lactate levels and ICU patient outcomes using the MIMIC II (version 2.6) database. We divided ICU admissions based on their initial lactate measurement into three groups: admissions with high lactate (above 4.0 mmol/L), admissions with medium lactate (between 2.0 mmol/L and 4.0 mmol/L), and admissions with low lactate (below 2.0 mmol/L). In addition to the ICU population as a whole, we studied sepsis patients using three different criteria (Martin, Angus, and infection with SIRS). We found that increased lactate levels were associated with a higher ICU mortality, higher 30 day mortality, longer ICU length of stay, and higher SOFA and SAPS I severity scores in all ICU admissions and in all three sepsis cohorts. Sepsis patients with high initial lactate levels were the most severely ill of all the patient populations. Sepsis patients identified with the Martin criteria who had high lactate levels had the worst outcomes of the three sepsis cohorts, but had similar average severity scores. This suggests that knowing lactate levels may give predictive value in addition to severity scores. We also investigated the relationship between initial lactate, change in lactate from the first measurement to the second measurement, and ICU mortality. We found that patients with high initial lactate levels in combination with an increase in lactate level typically had poorer outcomes than patients with high initial lactate levels with a decrease in lactate level.
by Hilary G. Mulholland.
M. Eng. in Computer Science and Molecular Biology

This study explores women's experiences of critical illness in Jordanian intensive care units. A narrative approach was employed to access Jordanian women's stories of their critical illness and to study how these accounts changed during the period following their discharge from intensive care. The study was conducted in two hospitals in a major Jordanian city. A purposive sample of 16 women who had spent at least 48 hours in intensive care was recruited over a period of six months, with each woman taking part in between one and three interviews during the six month period. Two focus group discussions were also conducted with 13 ICU nurses drawn from the hospitals in which the women had been patients. These had the aim of encouraging discussion about the development of new supportive care strategies for critically ill women in Jordanian intensive care units. The study findings revealed three main areas: the women's experiences of suffering and pain; their experience of body care; and the impact of the ICU experiences on their lives after discharge home. Experiences of suffering were pervaded with physical, emotional, social and temporal dimensions, interlinked with pain that was often severe, overwhelming, and disturbing to their sleep. The notion of 'nafsi' suffering was employed to describe emotional and social losses, such as loss of family support, which the women experienced. The notion of 'vicarious death' was used to explain the mortal fear women experienced in witnessing the death of others. Loss of body control, the unfamiliar ICU environment, and the sudden onset of illness made it difficult for the women to make sense of their experiences. This study shows that cultural norms and religious beliefs shape the ways in which these Muslim women made sense of their bodies. An analysis of the concept of 'bodywork' is presented: the 'dependent body' captures the women's experiences of changes of their physical status, which meant that from being care providers, they became those in need of care. This involved the experience of a sense of paralysis or disablement, and a complete dependence on their family or nurses. The 'social body' describes the women's feelings and emotions toward their family members. The latter assisted in the care of the women's bodies, but distress, frustration and a sense of loneliness were experienced by the women as a result of the loss of verbal communication with their relatives. The 'cultural body' describes the effect of cultural norms and Islamic religious beliefs on the women's interpretation of their experiences, and the interpretation of male nursing care in the ICU. The 'mechanical body' describes the women's experiences of the ICU machines as extensions of their bodies, and the senses of limbo and ambiguity they encountered during their ICU stay. The recovery period raised many physical, emotional, social, and spiritual issues, which in turn impacted on the women's experiences of their everyday lives. Weakness and tiredness accompanied with difficulties in eating and sleeping made some women feel frustrated and uncertain about their health. Some felt they were a burden upon their families. The meaning of the critical illness experiences were interpreted by some women as an opportunity to value family unity and neighbours' support. For other women, the illness experiences gave them lessons which strengthened their role as mothers and helped them to think positively about their future. This study highlights the importance of considering the cultural and religious preferences among Muslim women in critical care settings. The study recommendations focus on the need to base nursing care on an understanding of the physical, emotional, social, and religious elements of suffering, by exploring the potential of a palliative care approach for nursing critically ill people.

Electrical impedance tomography (EIT) is an emerging medical imaging technique that aims to reconstruct the internal conductivity distribution of a subject from electrical measurements obtained on the skin. In this thesis we explore the promising application of EIT to the respiratory monitoring of humans. We pay particular focus to the forward problem, highlighting the need to have an accurately known external boundary shape and electrode positions on a reconstruction model. A theoretical study of uniqueness results of EIT with an unknown external boundary shape is presented. A novel sensitivity study of the external boundary shape is presented as well as results from a reconstruction algorithm to account for errors in electrode position with simulated data in 3D. We also demonstrate results of a shape correction algorithm from a pilot study of lung EIT with data collected using the fEITER system, and MR images used to inform the external boundary shape of healthy subjects. After image co-registration of the resulting dynamic 3D EIT reconstruction images with the lung-segmented MR image, we outline a novel mutual information performance criterion to measure the quality of reconstructed images. We also outline the computation of the forward problem of the complete electrode model in 3D using high order polynomial finite elements and present convergence results in 2D for the continuum, point and complete electrode model. Our numerical study demonstrates that the convergence rate of the forward problem is independent of the polynomial approximation order for the complete electrode model and there is no global convergence for the point electrode model in the energy norm. Reconstructed conductivity images can be difficult to interpret at the bedside. Moreover clinicians would like clinically meaningful indices, such as regional lung compliance, to determine the pathologies of patients in real time. By modelling the respiratory system as a coupled time dependent system of simple mechanical functional units, we propose a novel methodology to couple mechanical ventilation and EIT. The mechanical properties of the lungs are estimated through an inverse coefficient problem on coupled ODEs, with the measurable data being the time series of pressure at airway opening and interior air volume data. We present results with simulated data as well as a discussion on extensions and limitations to the mechanical models. Finally we present a theoretical discussion of anisotropic EIT. It is well known that any diffeomorphism fixing points on the boundary gives rise to a conductivity with the same electrical measurements on the skin, generating a large class of conductivities that are electrically equivalent. We define novel classes of anisotropic media with constraints on their eigenspace: prescribed eigenvalues, prescribed orthogonal coordinates, prescribed eigenvectors, fibrous and layered conductivities. By drawing analogies with elasticity theory, we discuss how these constraints on the eigenspace restrict the set of diffeomorphisms fixing points on the boundary, and present two uniqueness results for anisotropic conductivities with prescribed eigenvalues and prescribed eigenvectors.

This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Thesis: S.M. in Technology and Policy, Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, 2019
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 121-130).
Patients in intensive care routinely have their blood glucose monitored and controlled using insulin. Two decades of on-going research has attempted to establish optimal glucose targets and treatment policy for patients with hyperglycemia in the intensive care unit (ICU). These efforts rely on the assumption that health care providers can reliably meet given targets. Significant proportions of the ICU population are either hypoglycemic or hyperglycemic and poor blood glucose control may lead to adverse patient outcomes. This thesis analyses approximately 20,000 ICU stays at the Beth Israel Deaconess Medical Center (BIDMC) which occurred between 2008 and 2018. These data are used to describe the state of clinical practice in the ICU and identify areas where treatment may be suboptimal. Even at a world-renowned teaching hospital, blood sugars are not optimally managed. 41.8% of diabetics and 14.2% of non-diabetics are severely hyperglycemic (>215mg/dL) each day. Insulin boluses are given more frequently than insulin infusions, despite guidelines recommending infusions for most critical care patients. When infusions are given, rates do not follow a consistent set of rules. Blood sugar management faces several challenges, including unreliable readings. Laboratory and fingerstick measurements that were taken at the same time had an R² of only 0.63 and the fingerstick measurements read on average 10mg/dL higher. Overcoming these challenges is an important part of improving care in the ICU. It is hoped that publicly sharing the code used to extract and clean data used for analysis will encourage further research. Code can be found at https://github.com/lawbaker/MIMIC-Glucose-Management
by Lawrence Baker.
S.M. in Technology and Policy
S.M.inTechnologyandPolicy Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society

The rapid digitalization of the healthcare domain in recent years highlighted the need for advanced predictive methods particularly based upon deep learning methods. Deep learning methods which are capable of dealing with time- series data have recently emerged in various fields such as natural language processing, machine translation, and the Intensive Care Unit (ICU). The recent applications of deep learning in ICU have increasingly received attention, and it has shown promising results for different clinical tasks; however, there is still a need for the benchmark models as far as a handful of public datasets are available in ICU. In this thesis, a novel benchmark model of four clinical tasks on a multi-center publicly available dataset is presented; we employed deep learning models to predict clinical studies. We believe this benchmark model can facilitate and accelerate the research in ICU by allowing other researchers to build on top of it. Moreover, we investigated the effectiveness of the proposed method to predict the risk of delirium in the varying observation and prediction windows, the variable ranking is provided to ease the implementation of a screening tool for helping caregivers at the bedside. Ultimately, an attention-based interpretable neural network is proposed to predict the outcome and rank the most influential variables in the model predictions’ outcome. Our experimental findings show the effectiveness of the proposed approaches in improving the application of deep learning models in daily ICU practice.

End-of-life care (EOLC) has become a significant area of expertise in the intensive care unit (ICU). Critical care nurses are the primary caregivers of patients in the ICU and they provide EOLC for patients and families daily. Nurses have portrayed EOLC as difficult and demanding work; yet, they have also described their experiences of providing EOLC as rewarding, gratifying, and a privilege. The purpose of this study was to explore nurses’ meaningful experiences with providing EOLC for patients and families in the context of the ICU. Van Manen’s approach to interpretive phenomenology was used. Unstructured face-to-face interviews were conducted with six registered nurses who were employed in a medical/surgical tertiary care ICU. The interviews were audio-recorded, transcribed, and analyzed. The essence of nurses’ meaningful experiences in providing EOLC was ‘being able to make a difference’. For the nurses, being able to make a difference reflected their efforts to create a good death for the dying patient and their family. The nurses had to navigate a variety of challenges that affected the creation of a good death, however, they made it work by building relationships quickly with families, taking care of themselves, and recognizing it’s a privilege to provide EOLC. These research findings contribute to an expanding body of knowledge and understanding with regards to nurses’ role with the provision of EOLC in the ICU.

Objective: The Patient Self-Determination Act of 1990 (PSDA) protects a patient’s right to predetermine the level of life-supporting care they are willing to receive (U.S. Department of Health and Human Services, 1993). In Arizona, the advance directive (AD) complies with the PSDA and is used to guide care in the event of cardiopulmonary failure. The AD may indicate “do not resuscitate” (DNR), which prohibits cardiopulmonary resuscitation in the event of cardiac arrest. In the institution used for this project, a palliative care team assists with identifying goals of care and helps guide interventions consistent with the AD. The purpose of this Doctor of Nursing Practice (DNP) project was to complete a retrospective chart review and identify patients admitted to the medical intensive care unit (ICU) with a DNR as indicated by a copy of the AD in the electronic health record (EHR) and determine if they received a palliative care consultation. This information could support a quality improvement project led by the DNP-prepared AGACNP focused on ensuring a palliative care consultation within 48 hours of admission for patients admitted to the ICU with a DNR. Methods: A search of the EHR identified patients admitted to the medical ICU over a 12-month period. The EHR of patients admitted with a DNR were reviewed to determine if they received a palliative care consultation during the ICU stay and the patient’s final disposition. Findings: A total of 38 patients had an AD indicating DNR status on admission to the medical ICU. Of those patients, 26 (68.4%) received a palliative care consultation. Twelve patients (31.6%) with a DNR status on admission did not receive a palliative care consultation. Additionally, five patients with a DNR (13.16%) died in the ICU without receiving a palliative care consultation. Conclusion: Twelve patients with an AD indicating a DNR did not receive a palliative care consultation, and five of those patients died in the ICU. The findings from this project support a quality improvement project to implement palliative care consults to review goals of care for patients with a pre-existing AD indicating a DNR code status.

CoordenaÃÃo de AperfeÃoamento de Pessoal de NÃvel Superior
Hospitalization of the newborn is necessary when health conditions require immediate assistance for their recovery. Humanized actions in the neonatal unit have been developed in order to make it less painful separation parent-child when it needs technological support and team of trained professionals. It was aimed to analyze the perception and knowledge of the nursing team on the promotion of humanized care for newborn in a Neonatal Intensive Care Unit . It is a qualitative study conducted in a public hospital, large, tertiary level, in Fortaleza, Brazil, in the months October and November 2015, after approval by the Research Ethics Committee, under Protocol N. 1,191,339. The subjects were 14 nurses and 20 nursing technicians working in neonatal care. The data collected through semi-structured interviews consist identification data and five guiding issues that permeate the knowledge of the nursing team about the care and promotion of humanized care in the UTIN. In addition, we used no-participant observation and field diary. For analysis, we sought to Bardin technique that extracted the three categories lines: âTaking care of the human personâ, ânursing contributions to the humane careâ and âFactors that affect the quality of humanized care.â The results showed that the nursing team understands humanization as an indispensable element for the comprehensive care to the baby and family, which was observed from the speeches of welcome, restoring health and disease of the newborn process. The professionals had knowledge of the humanized care, played their actions conscious, oriented and appreciative way about the quality of neonatal care and parents who face the challenges inherent in the admission process. We conclude that the performance of these professionals permeates compliance with the regulations of the National Humanization Policy regarding humanized care to the newborn, family and neonatal ambience. It is believed that such actions minimize the impact caused by the characteristics of the disease treatment as well as stressors.
A hospitalizaÃÃo do recÃm-nascido faz-se necessÃria, quando as condiÃÃes de saÃde requerem assistÃncia imediata para o seu restabelecimento. As aÃÃes humanizadas na unidade neonatal tÃm sido desenvolvidas, a fim de tornar menos dolorosa à separaÃÃo pais-filho, quando este necessita de suporte tecnolÃgico e equipe de profissionais capacitados. Objetivou-se analisar a percepÃÃo e conhecimentos da equipe de enfermagem sobre a promoÃÃo do cuidado humanizado ao recÃm-nascido internado na Unidade de Terapia Intensiva Neonatal (UTIN). Trata-se de estudo qualitativo, realizado em hospital pÃblico, de grande porte, nÃvel terciÃrio, em Fortaleza-CE-Brasil, nos meses outubro e novembro de 2015, apÃs aprovaÃÃo pelo Comità de Ãtica em Pesquisa, sob Protocolo n 1.191.339. Os sujeitos foram 14 enfermeiros e 20 tÃcnicos de enfermagem atuantes na assistÃncia ao neonato. Os dados coletados, por meio de entrevista semiestruturada, consistem dados de identificaÃÃo e cinco questÃes norteadoras, que permeiam o conhecimento da equipe de enfermagem acerca do cuidado e a promoÃÃo da assistÃncia humanizada na UTIN. Ademais, utilizou-se observaÃÃo nÃo participante e diÃrio de campo. Para anÃlise, sÃntese e descriÃÃo, buscou-se a tÃcnica de Bardin, que se extraÃram das falas trÃs categorias: âCuidar do ser humanoâ, âContribuiÃÃes de enfermagem para o cuidado humanizadoâ e âFatores que interferem na qualidade do cuidado humanizadoâ. Os resultados revelaram que a equipe de enfermagem compreende a humanizaÃÃo como elemento indispensÃvel para o cuidado integral ao bebà e famÃlia, o que se observou desde as intervenÃÃes de acolhimento, ao restabelecimento do processo saÃde-doenÃa do neonato. Os profissionais apresentaram conhecimentos acerca do cuidado humanizado, desempenharam suas aÃÃes de forma consciente, orientada e sensibilizada, quanto à qualidade da assistÃncia ao neonato e aos pais que enfrentam os desafios inerentes ao processo de internaÃÃo. Percebe-se, portanto, que a atuaÃÃo desses profissionais permeia o cumprimento aos regulamentos da PolÃtica Nacional de HumanizaÃÃo. Conclui-se que o cuidado humanizado aplicado nessa ambiÃncia à essencial ao recÃm-nascido e famÃlia, uma vez que minimiza o impacto causado pelas caracterÃsticas da doenÃa, tratamento, bem como os fatores estressantes da UTIN.