Weebly



Assignment: State of Science Paper TwoTopic: Therapeutic Hypothermia to Improve Neurological and Survival Outcome in Adult Post Cardiac Arrest Patients: A Systematic ReviewRoshan Jan Muhammad The Johns Hopkins University School of Nursing “On my honor, I pledge that I have neither given or nor received any unauthorized assistance on this assignment”. RJMAbstractCardiac arrest (CA) and related mortality is one of the major health concerns in the United States. The most evident cause of death among post CA patients is anoxic brain injury that accounts for two third mortalities in these victims (Stub et al., 2011). The mechanism of brain injury is triggered by ischemia during cardiac arrest and is further aggravated during reperfusion phase when patient has return of spontaneous circulation (ROSC). Along with other measures, mild therapeutic hypothermia (TH) is also proposed by International Liaison Committee of Resuscitation (ILCOR), in order to prevent neurologic adversity and related mortality in CA patients (Peberdy et al., 2010). However, despite international guidelines directing the management of post CA patients, TH is used in less than 1% of CA patients in USA hospitals because of statistical uncertainty of research findings (Kim et al. 2012). The purpose of this state of the science paper is to determine the effect of TH on neurological and mortality outcome of adult post CA patients compare to normo-thermia, in order to establish the basis for nursing practice. Out of seventeen relevant articles found for the period of 2007-2012, ten were reviewed for this paper that included; two Randomized Controlled Trial (RCT); one Quasi-Experiment study; six retrospective observational studies and one prospective study. It is synthesized from individual evidence summaries that here is evidence supporting the beneficial effect of therapeutic hypothermia in reducing mortality and poor neurological outcomes in adult post CA patients. However, much of these evidences are from observational studies (level III); in many studies positive outcomes fail to reach statistical significance; and most studies in the review had substantial risk of bias and had compromised validity. Therefore, a rigorous RCT is proposed as nursing implication. Key words: Therapeutic hypothermia (TH),Cardiac arrest (CA), return of spontaneous circulation (ROSC), shockable rhythm (ventricular fibrillation or pulseless ventricular tachycardia), non-shockable rhythm (Asystole or pulseless electrical activity), anoxic brain injury, neurological outcome, mortality.IntroductionAccording to American Heart Association (AHA), cardiac arrest (CA) is defined as “an abrupt loss of heart function that is confirmed by the absence of circulation” (Beddingfield et al., 2012). It is reported that out of hospital cardiac arrest (OHCA) affects up to 236,000 to 325,000 people in United States each year, with average survival to discharge rate of 7% (Roger et al., 2011). On the other hand, In-hospital CA mortality rate of 67% is also a staggering figure (Neumer et al., 2008). Though cardiac arrest causes global ischemia, but the brain is most vulnerable to metabolic failure secondary to circulatory collapse during CA. According to Neumer et al. (2008), neurological damage starts in 4-6 minutes after sudden cardiac arrest if resuscitation is not attempted. However, immediate cardiopulmonary resuscitation and early defibrillation increases the probability of return of spontaneous circulation (ROSC) and perfusion. There is variability in definition of term “ROSC” in the literature. Nevertheless, it is referred as sustained return of pulse or spontaneous circulation for the period ranging from 30 seconds to 20 minutes (Neumer et al., 2008). While, ischemia is detrimental to the brain, evidences also suggest that reperfusion after a brief episode of global ischemia, compliments initial hypoxic insult to the body, by causing mitochondrial dysfunction. Pathophysiological responses during the early period of reperfusion include calcium over load, abundance of reactive oxygen species and abrupt restoration of PH. As a result, opening of non-specific mitochondrial permeability transition pore (mPTP) is stimulated, leading to cell death (Cour et al., 2011). It is also known that ischemic event perpetuate release of pro-inflammatory cytokines as a cascade of cellular events, thence, compounding injury to the brain cells. Consequently, those who survive CA suffer serious debilitating neurological complications or die due to post cardiac arrest syndrome. Nolan et al. (2010) further emphasizes that the risk of poor neurological outcome post ROSC increases with each degree increment in body temperature above 37C. Mild therapeutic hypothermia (TH), that ranges between 32C-34C, is known to improve neurological outcomes of post CA patients through various neuro-protective effects (Nolan et al., 2010). TH prevents astroglial proliferation and blocks the cascade of pro inflammatory mediators. It mitigates brain damage by reducing cerebral metabolic requirement; decreasing cerebral edema, and intracranial pressure; inhibiting reperfusion injury; and limiting apoptosis (Beddingfield et al., 2012). TH is a tri-phasic intervention which ideally include; induction, maintenance and rewarming phase. Typically, induction phase is executed using conventional ice packs on the groin, in armpit, and around the neck; intravenous ice-cold 0.9% normal saline or ringers lactate infusion; and /or contemporary surface or internal cooling devices. This phase often involves concomitant administration of opioids, hypnotics or neuromuscular blocking agents to avoid shivering, as it has potential to aggravate neurological damage. Maintenance phase usually lasts for 12-24 hours. Lastly, rewarming is done by using internal or external thermal devices to ensure controlled rewarming at rate of 0.25C to 0.5C per hour (Nolan et al., 2010). For the first time, neuro-protective effects of TH in post CA patients got validation in 2002, when two independent randomized controlled trial by Bernard et al. (2002) and HACA study group (2002) presented positive outcomes of TH in OHCA cases with an initial rhythm of ventricular fibrillation (VF). Subsequently, in 2003, International Liaison Committee of Resuscitation (ILCOR) recommended the use of TH in the treatment of adult post cardiac arrest patients (Nolan et al., 2003). However, the recommendation was constraint to those CA victims whose initial rhythm was shockable. Later, subsequent studies further supported the findings. Nonetheless, most of them were observational. In 2010, scientific statement released by American Heart Association on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care, carried on with same recommendation (Nolan et al., 2010). This time, the recommendation expanded the scope to patients who sustained CA in-hospital and patients who had initial non-shockable rhythm. Bernard (2012) reports that incidence of CA with PEA and Asystole as an initial rhythm is rising. He further adds that, contrary to CA patient with shockable rhythm, this cohort has a relatively poor prognosis. Meta-analysis conducted by Kim et al. (2012) also concluded that, though TH is associated with reduced mortality for adult patients resuscitated from non-shockable rhythm, most studies in the review had substantial risk of bias and the quality of evidence was very low. Therefore, despite recommendations from expert panel, utilization of TH is still limited and is only used for less than 1% of CA patients in USA hospitals (Kim et al. 2012). Major barriers to wide spread use of TH are skepticism about quality of evidences, uncertainty about eligibility, doubts about long term benefit, misconception of futility and lack of motivation. The purpose of this state of the science paper was to determine if therapeutic hypothermia compared with normo-thermia improves mortality and neurological outcomes in adult post cardiac arrest patients with ROSC. In my future role of clinical nurse specialist, I envision myself as a change agent advocating for evidence based practice. This review assisted me critically appraise related evidences and synthesize conclusion about this crucial area of clinical practice. The paper explored evidences pertinent to TH intervention in patients with both non-shockable and shockable rhythm.Search engines like PubMed, EMBASE, CINAHL and Cochrane were explored for the period of 2007-2012 during search process. Database specific MeSH terms were used using Boolean Operators. The search terms included; for PubMed ((“heart arrest”) OR “cardiopulmonary resuscitation”) AND “hypothermia, induced”; for CINAHL (MM “hypothermia, induced”) AND (MM “heart arrest”); for EMBASE ‘heart arrest’/exp OR ‘resuscitation’/exp AND ‘induced hypothermia’/exp; and for Cochrane “hypothermia”. The search was truncated to randomized control trials, quasi-experimental studies and non-experimental quantitative studies conducted on humans having ages above 18years. Publications in language other than ‘English’ were excluded. The search retrieved total 271 articles, following breakdown of PubMed (42), CINAHL (10), Cochrane (1), and EMBASE (218). After eliminating duplication through ref works, the abstracts of all articles were reviewed based on inclusion criteria. Finally, seventeen articles were found relevant, of which, ten were used for this state of science paper. The following summary is synthesized from individual evidence evaluation of all ten articles (Appendix B).Overall Summary of Evidences Studies DescriptionsOf ten articles appraised, two included RCTs (Tiainen et al., 2007; Kim et al., 2007); one was quasi-experimental study (Granja et al., 2011); six were retrospective observational studies (Reinikainen et al., 2012; Prior et al., 2010; Pfeifer et al., 2011; Testori et al., 2011; Stub et al., 2011; Van der et al., 2011); and one had prospective observational design (Storm et al., 2012). All were of good quality except for Kim et al.’s (2007) study that was rated poor. All the studies included hypothermia in the treatment arm; however, there was variability in control/comparison groups. Amongst ten studies, two had control group of normo-thermia with controlled hyperthermia (Tiainen et al., 2007; Pfeifer et al., 2011); two utilized normo-thermia group without hyperthermia control (Testori et al., 2011; Van der et al., 2011); whereas, five studies (Granja et al., 2011; Reinikainen et al., 2012; Prior et al., 2010; Storm et al., 2012; Stub et al., 2011) did not provided information about hyperthermia control in comparison group. All observational studies had used historical controls for the comparison. Overall, these studies enrolled 10,912 patients, 7472 in the treatment arm and 3396 in the comparison group, out of which, 9275 patients were in two registry based multicenter-retrospective observational studies only. Out of these ten studies, three studies inducted CA patients with shockable rhythms; two included CA patients from non-shockable rhythms; and remaining five studies enrolled CA survivors from both shockable and non-shockable rhythms. Remaining key facts about the study design and the results are summarized in Appendix A. Effect of Therapeutic Hypothermia on Survival Outcomes Post Cardiac Arrest Out of ten studies, nine showed mortality benefit of TH in post CA patients. However, findings of only five studies (Reinikainen et al., 2012; Prior et al., 2010; Testori et al., 2011; Stub et al., 2011; Van der et al., 2011) could reach statistical significance, and all of them were observational studies with historical controls. In addition, all statistically significant results were observed against normo-thermia group in which hyperthermia was either not prevented or not monitored/reported. Discharge outcomes of intervention group in CA patients with shockable versus non-shockable rhythm is also important to analyze. All three studies (Tiainen et al., 2007; Reinikainen et al., 2012; Stub et al., 2011) with inclusion of CA patients with shockable rhythm had better survival rate at discharge, and two of them reached statistically significant conclusion as well. On the contrary, two studies (Testori et al., 2011; Storm et al., 2012) that had enrolled CA patients with non-shockable rhythm, only one showed beneficial effect and significant results. Synthesis of remaining four studies (Granja et al., 2011; Prior et al., 2010; Pfeifer et al., 2011; Van der et al., 2011) that had inducted both shockable and non-shockable rhythms is equivocal. Though all of them reported reduced mortality rate, nonetheless, only two had significant results. Of those two, only one that presented separate analysis concluded substantial favorable outcome in patients with shockable rhythm versus non-shockable rhythm. Overall evidences, though of lower quality, are more favorable for CA victims with shockable rhythms. Nevertheless, clinical benefits of non-statistically significant, yet inconsistent positive results cannot be disregarded for the non-shockable group as well. However, in lieu of threats to internal validity, that is described latter in the paper, findings should be generalized with great caution. Effect of Therapeutic Hypothermia on Neurological Outcomes Post Cardiac Arrest Of eight studies that evaluated neurological status post treatment, all endorsed protective neurological effect of TH in the treatment group. Nevertheless, only four could render statistically significant conclusion (Prior et al., 2010; Pfeifer et al., 2011; Testori et al., 2011; Stub et al., 2011), all being lower level of evidence (LOE III) and with historical controls. It is important to note that three out of four studies had comparable group with no pyrexia preventive measures. Separate analysis of the findings pertinent to shockable and non-shockable rhythms is also critical. TH attributed to positive outcomes in both studies done on CA patients following shockable rhythm, but, only one produced statistically conclusive results. Similar outcome pattern emerged from two studies that included CA patients with non-shockable rhythm. Conversely, four studies that enrolled CA patients with all rhythms, despite positive outcomes in all, only two produced statistically substantial results, mostly in VF/VT patients. LimitationsAll the studies included in the analysis had vast differences in terms of study deign, rigor, and methodology. Therefore, aforementioned, results must be viewed in consideration of study’s limitations. Selection bias, lack of control over confounding variables, inadequate standardization of treatment, sample size limitation and influence of history as a threat to internal, external and statistical validity of the studies are discussed here to guide conclusion.Selection BiasOf ten studies reviewed, five included the samples having a particular type of rhythm in CA patients. Additionally, non-random sampling, exclusion of critical patients, elimination of fatality cases within 24 or 48 hours of CA from the study, and the use of subgroup for the analysis, are few critical elements that contribute to selection bias. As a consequence, study sample may not be true representative of all CA patients. Therefore, the findings cannot be generalized to all CA patients. Lack of Control over Confounding VariablesOverall the studies reviewed have presented suboptimal control over following critical confounding variables that compromises internal validity of the studies.Fever is common after CA. It is responsible for release of excitatory neurotransmitters that is responsible for neuronal cell death, and unfavorable neurologic outcome in a variety of experimental models of brain injury (Laptook et al., 2008). Natale, et al. (2000) documents that fever in CA victim is independently associated with worse outcome compared to patients with normothermia. In this systematic review, we found that only Tiainen et al. (2007) and Pfeifer et al. (2011) had controlled hyperthermia in their comparison group and both the studies showed positive but not statistically significant outcomes in TH group. It is interesting to note that in Van der et al’s. (2011) study, median temperature in the treatment arm was inclined toward lower threshold of normo-thermia (35.5C) in most patients instead of hypothermia, and that study also showed positive yet statistically inconclusive results. Thus, it can be extrapolated that lower threshold normo-thermia with fever control measures may produce the comparable results as of TH. It is consistent with the conclusion of Stub et al.’s (2011) study that pyrexia prevention can be used as a minimum recommendation if TH is contraindicated or not feasible. Arterial Hyperoxia (PaO2 >300mmHg) is another potential confounder that is labeled as a predicator of mortality in patients with CA compared with normoxia and hypoxia (Kilgannon et al., 2010). Additionally, Mean arterial pressure (MAP) is also listed as an important covariate for poor outcomes in CA victims. Stub (2011) reports that hemodynamic instability (MAP < 65mmHg), that is common after effect of CA, is also associated with poorer survival and neurological outcomes. Furthermore, Nielson et al. (2011) has reported that sustained hyperglycemia (blood glucose >8 mmol/L for >4 hrs) is associated with high mortality in post CA patients who were treated with TH. During this review, it was found that only Tiainen et al. (2007); Kim et al. (2007); and Prior et al. (2010), demonstrated reasonable control over these variables through exclusion or regression analysis. Though all three studies have shown positive impact but, only Prior et al. (2010) has documented significant outcomes. It is plausible that poor outcomes in normo-thermia group are secondary to lack of control over this extraneous variable.In recent years, TH with Percutaneous Coronary Intervention (PCI) is also surfacing as efficacious approach to attain better outcomes in comatose survivors of OHCA. Stub et al’s. (2011) study used this synergy intervention and concluded statistically significant results. But, Stub did not present multivariate analysis to statistically adjust the effect modification or confounding effect of PCI on the association of TH and outcomes.Overall, except for Tiainen et al. (2007), no other studies have presented adequate control or statistical adjustment of above mentioned covariates and his research also revealed non- statistically significant positive results. Inadequate Measures to Standardize Treatment All observational studies in particular, by virtue of their design, had huge variability in treatment implementation. Burns and Groove (2009) emphasizes that treatment standardization is crucial every time it is executed to detect a true difference. Inconsistency in treatment implementation reduces the likelihood of detecting the true difference. In studies reviewed, researchers did not report measures of central tendency and variability for induction time for TH (Tiainen et al., 2007 ; Kim et al., 2007; Reinikainen et al., 2012; Storm et al., 2012; Stub et al., 2011; Van der et al., 2011); temperature achieved against target goal in TH group (Kim et al., 2007 ; Granja et al., 2011; Reinikainen et al., 2012; Prior et al., 2010; Storm et al., 2012; Stub et al., 2011); duration of cooling (Kim et al., 2007; Reinikainen et al., 2012; Pfeifer et al., 2011; Testori et al., 2011; Storm et al., 2012; Stub et al., 2011;Van der et al., 2011); and rewarming rate. Nordmark et al. (2009) reports that neurochemical changes indicating cerebral ischemia and excitoxicity are found both after CA and during the re-warming phase that perpetuate neurological sequel. Thus, delayed induction and rapid rewarming might decrease the neuro-protective effects of TH. Due to gaps in information, best parameters of TH execution cannot be established. Instrument and Measurement IssuesCore temperature monitoring was of essence in the study. Reinikainen et al. (2012); Stub et al. (2011) and Van der et al. (2011) did not report the monitoring instrument used in their studies. Nonetheless, Kim et al. (2007) and Testori et al. (2011) have reported inconsistent use of instrument that might have compromised the precision of measurements. One the other hand, instrument used for neurological evaluation, and it methodology is critical to appraise aforementioned presented synthesis. Though all studies used Pittsburgh Cerebral Perfusion Scale, Prior et al. (2010); Pfeifer et al. (2011); Storm et al. (2012); and Stub et al. (2011) assigned ratings to the patients through clinical chart review instead of direct evaluation. The remaining researchers, who used in person evaluation, only Tiainen et al. (2007) and Testori et al. (2011) used blinded assessor to prevent bias. Additionally, time to assessment was also variable among all the studies. All researchers have presented neurological outcomes at discharge except for Tiainen and Testori, who did appraisal at 3 month and 6 months. As a consequence, patients who might have recovered in long run, could not be captured in all remaining studies that might have under-estimated the outcomes.Statistical validity According to Burns and Groves (2009), low powered studies increases the likelihood of concluding non-statistically significant result between groups when actually there is difference. Despite the fact, only Reinikainen et al. (2012) and Van der et al.’s (2011) have used large sample size, concluding non-significant and significant findings respectively, causing ambivalence. Besides that, none of the studies, including level I and II trails, reported power analysis thus, adequacy of sample size cannot be ascertained and non-statistically significant findings remain inconclusive and dubious. History as a Threat to Validity Peri-cardiac arrest factors like duration and quality of BLS and ACLS and general post resuscitation care are important determinants of CA outcomes. Revised cardiopulmonary resuscitation guidelines published from ILCOR, in 2003, was a break through. It recommended several key changes to reduce no flow (time from collapse to BLS) and low flow time (time from collapse to ROSC); and ameliorate post ROSC care. It can be deduce that promising results in CA victims after this period can be due to bundle effect of these recommendations. And therefore, significant results of Reinikainen et al. (2012); Prior et al. (2010); Testori et al.’s (2011); Storm et al.’s (2012); Stub et al.’s (2011); Van der et al.’s (2011) studies, that utilized historical controls from before this period can be challenged by rival hypothesis.ConclusionThis state of the science paper concludes that there is reasonable evidence that TH is effective for adult comatose post CA victim with ROSC. Nevertheless, favorable evidences are more inclined towards CA with shockable rhythm and all studies with statistically significant conclusion are of lower level and lack adequate rigor. Therefore, causality cannot be established and TH can be concluded as a sole and independent predictor of positive outcome in CA patients. However, keeping in view that outcomes in CA patients are generally very poor, non-statistically significant survival and neurological benefits might be of clinical importance. Despite this consideration, clinical implication of TH in this patient population is highly restricted. It is because, there are several questions still unanswered due to study limitations and in-adequate information, such as, who would be the beneficiaries; what is the best induction time and duration of cooling, what is most therapeutic depth of cooling and rate of rewarming. As a consequence, it limits translation of this research synthesis into practice. Therefore, Randomized Control Trial with adequate rigor and with better control over confounding variables needs to be conducted, that draws more definite conclusion, determines causal relationship, produces unbiased results for patients with both shockable and non-shockable rhythms, and adequately answers the key questions related to treatment implementation. Besides, that also proclaims the long term survival and neurological benefit of the treatment. Thus, based on the learning from this synthesis, following study is proposed for future nursing implication. Implications for Nursing Research Research Question To evaluate if therapeutic hypothermia (32-34C) has positive survival and neurological benefit compare to lower threshold normo-thermia(36C) in adult post CA patients with shockable and non-shockable rhythm.Study Design Multicenter randomized Controlled trial. Setting. ICU settings of at least 3 tertiary care hospitals would be used to conduct the study. Variables Independent variables. Hypothermia and lower threshold normo-thermia would be independent variables in the study. Dependent variables. There are 2 dependent variables in the study. First is ‘survival outcome’ that would be monitored at discharge, one month, three months and six months period through hospital data base and follow up through telephone by blinded assessor. Second variable is ‘neurological outcome’, which would also be followed at the same duration as of survival outcome. Evaluation would be carried out by blinded assessor through in person evaluation of patients. We would use 2 valid and reliable instruments to prevent mono-opertaion bias that include: Pittsburgh Cerebral Perfusion Category (CPC) and Modified Rankin Scale (MRS). Standard definition would be used to dichotomize the neurological outcome as ‘good’, that is CPC < 2 and MRS < 3.Inclusion Criteria All patients over 18 years of age, who are resuscitated from OHCA or in-hospital CA with presumed cardiac causes and who remain unconscious (GCS <8) one hour after ROSC would be included in the study. It is aligned with landmark RCT’s (HACA, 2002 & Bernard et al., 2002) and Tiainen et al.’s (2007) clinical trial. Exclusion Criteria We would follow the exclusion criteria similar to most of the RCTs reported so far. The criteria include pregnancy; coagulopathy; intracranial hemorrhage and stroke; CPC > 2; cardiogenic shock (i.e., MAP systolic blood pressure less than 80 mmHg) refractory to pressors; and terminal disease before cardiac arrest.Sample Size and Sampling Methodology To ensure that study is adequately powered to detect the difference in the outcomes between groups, sample size would be calculated to attain 80% power at 5% level of significance. However, effect size would be extrapolated from meta-analysis of high or good quality randomized control trial on related theme. As we intend to conduct neurological evaluation at discharge, 3 months and 6 months, potential sample attrition would also be considered while calculating the sample size. Patients would be randomly assigned to treatment and controlled arm. Nature of the study is such that the care providers cannot be kept blinded, nevertheless, person conducting neurological evaluation and the statistician would be blinded. Method The hypothermia protocol would be developed following recommendations from American Association and ILCOR to ensure standardization of treatment. Prior to the start of therapy, baseline information would be obtained that is peri-cardiac arrest, baseline vital signs, lab investigation, clinical evaluation and acute physiology and chronic health evaluation (APACHE) score to be used later for multivariate analysis. Thereafter, TH would be introduced using intravenous cold saline and temperature would be monitored via urinary catheter. The intervention would be phased out in 4 steps. (a)‘Induction phase’ would be within 4 hours of ROSC. (b) ‘Maintenance phase’ would last for 24 hours from individual patients’ indication time after achieving temperature end point. (c) ‘Rewarming phase’ would have end point of 37 C with rate adjusted for 8 hours duration. (d) ‘Post hypothermia phase’ would include maintenance of temperature 37 C + 0.5 C to prevent detrimental effect of pyrexia for 72 hours from CA in both the groups. All confounding variables listed in the limitation section would be monitored throughout 72 hours.In consideration of presumed certainty about deleterious effect of hyperthermia, temperature for control group is determined as lower threshold normo-thermia to prevent potential sweep to hyper thermic range. In order to do so, close monitoring and aggressive quality control measures would be undertaken to maintain temperature at cut off range and to prohibit hyperthermia in control group. Besides that, both the groups would receive comparable treatment for shivering and seizures prevention; glycemic control (i.e., 8-10 mmol/L); MAP maintenance (above 65 mmHg, through fluids and vassopressors); and hypoxia/hyperxia prevention. This would control key confounding variables related to treatment that were poorly controlled in previous studies. Statistical Analysis. Analysis would essentially include comparison of the groups for basic characteristics using t test for continuous variable and Chi-square for categorical variables. Main outcome analysis would be pursued as follows.‘Survival outcome at discharge’ would be carried out by using uni-variate logistic regression. To eliminate the potential effect of confounder or effect modifier, multi-variate logistic regression will be performed, adjusted for covariates like no flow time, low flow time, initial rhythm, blood glucose level, arterial oxygen level, and MAP, pyrexia, gender, age and acuity level based on (APACHE II) score. Additionally, ‘survival outcome at 6 months’ would be determined using cox regression analysis both unadjusted and adjusted for above listed variables. ‘Neurological outcome at discharge, 3 months and 6 months’ would be evaluated through Chi- square to compare proportion of good and bad neurological outcomes between the groups. Post hoc stratified analysis will also be conducted to present the outcomes based on shockbale and non-shockable rhythm of cardiac arrest. For all the tests, the effect estimates will be provided considering 95% confidence interval and ∝ 0.5 (2 tailed).Stake Holders The nature of the study demands involvement of multidisciplinary team that include administration of hospital CPR committee; chair of ICU committee; neurologist; nurse manager of the units; clinical nurse specialists; nurses and medical residents on the floor; patients and their proxy; and statistician . We would also involve experts in the field from ILCOR to provide consultancy. ReferencesBeddingfield, E., & Clark, A. P. (2012). Therapeutic hypothermia after cardiac arrest: Improving adherence to national guidelines. Clinical Nurse Specialist CNS, 26(1), 12-18. doi:10.1097/NUR.0b013e31823f8a02 Bernard, S. A., Smith, K., Cameron, P., Masci, K., Taylor, D. M., Cooper, D. J., . . . Rapid Infusion of Cold Hartmanns Investigators. (2012). Induction of prehospital therapeutic hypothermia after resuscitation from nonventricular fibrillation cardiac arrest*. Critical Care Medicine, 40(3), 747-753. doi:10.1097/CCM.0b013e3182377038 Bernard, S. A., Gray, T. W., Buist, M. D., Jones, B. M., Silvester, W., Gutteridge, G., & Smith, K. (2002). Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. The New England Journal of Medicine, 346(8), 557-563. doi:10.1056/NEJMoa003289 Burns, N., & Grove, S. K. (2009). The practice of nursing research : Appraisal, synthesis, and generation of evidence (6th ed.). St. Louis, Mo.: Saunders/Elsevier. Cour, M., Loufouat, J., Paillard, M., Augeul, L., Goudable, J., Ovize, M., & Argaud, L. (2011). Inhibition of mitochondrial permeability transition to prevent the post-cardiac arrest syndrome: A pre-clinical study. European Heart Journal, 32(2), 226-235. doi:10.1093/eurheartj/ehq112 Cullen, D., Augenstine, D., Kaper, L., Tinkham, S., & Utz, D. (2011). Therapeutic hypothermia initiated in the pre-hospital setting: A meta-analysis. Advanced Emergency Nursing Journal, 33(4), 314-321. doi:10.1097/TME.0b013e3182343cb6 Granja, C., Ferreira, P., Ribeiro, O., & Pina, J. (2011). Improved survival with therapeutic hypothermia after cardiac arrest with cold saline and surfacing cooling: Keep it simple. Emergency Medicine International, 2011, 395813. doi: 10.1155/2011/395813 Hypothermia after Cardiac Arrest Study Group (HACA). (2002). Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. The New England Journal of Medicine, 346(8), 549-556. doi:10.1056/NEJMoa012689 Kilgannon, J. H., Jones, A. E., Shapiro, N. I. (2010). Association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality. JAMA, 303, 2165-2171.Kim, F., Olsufka, M., Longstreth, W. T.,Jr, Maynard, C., Carlbom, D., Deem, S., . . . Cobb, L. A. (2007). Pilot randomized clinical trial of prehospital induction of mild hypothermia in out-of-hospital cardiac arrest patients with a rapid infusion of 4 degrees C normal saline. Circulation, 115(24), 3064-3070. doi: 10.1161/CIRCULATIONAHA.106.655480 Kim, Y. M., Yim, H. W., Jeong, S. H., Klem, M. L., & Callaway, C. W. (2012). Does therapeutic hypothermia benefit adult cardiac arrest patients presenting with non-shockable initial rhythms?: A systematic review and meta-analysis of randomized and non-randomized studies. Resuscitation, 83(2), 188-196. doi:10.1016/j.resuscitation.2011.07.031 Laptook A, Tyson J, Shankaran S, et al.: Elevated temperature after hypoxic-ischemic encephalopathy: risk factor for adverse outcomes. Pediatrics 2008; 122:491–499Natale JE, Joseph JG, Helfaer MA, et al.: Early hyperthermia after traumatic brain injury in children: Risk factors, influence on length of stay, and effect on short-term neurologic status. Crit Care Med 2000; 28:2608–2615Neumar, R. W., Nolan, J. P., Adrie, C., Aibiki, M., Berg, R. A., Bottiger, B. W., . . . Vanden Hoek, T. (2008). Post-cardiac arrest syndrome: Epidemiology, pathophysiology, treatment, and prognostication. A consensus statement from the international liaison committee on resuscitation (american heart association, australian and new zealand council on resuscitation, european resuscitation council, heart and stroke foundation of canada, InterAmerican heart foundation, resuscitation council of asia, and the resuscitation council of southern africa); the american heart association emergency cardiovascular care committee; the council on cardiovascular surgery and anesthesia; the council on cardiopulmonary, perioperative, and critical care; the council on clinical cardiology; and the stroke council. Circulation, 118(23), 2452-2483. doi:10.1161/CIRCULATIONAHA.108.190652 Nolan, J. P., Morley, P. T., Vanden Hoek, T. L., Hickey, R. W., Kloeck, W. G., Billi, J., . . . International Liaison Committee on Resuscitation. (2003). Therapeutic hypothermia after cardiac arrest: An advisory statement by the advanced life support task force of the international liaison committee on resuscitation. Circulation, 108(1), 118-121. doi:10.1161/01.CIR.0000079019.02601.90 Nolan, J. P., Neumar, R. W., Adrie, C., Aibiki, M., Berg, R. A., Bbttiger, B. W., . . . Council on Stroke. (2010). Post-cardiac arrest syndrome: Epidemiology, pathophysiology, treatment, and prognostication: A scientific statement from the international liaison committee on resuscitation; the american heart association emergency cardiovascular care committee; the council on cardiovascular surgery and anesthesia; the council on cardiopulmonary, perioperative, and critical care; the council on clinical cardiology; the council on stroke (part II). International Emergency Nursing, 18(1), 8-28. doi:10.1016/j.ienj.2009.07.001 Nordmark, J., Rubertsson, S., Mortberg, E., Nilsson, P., & Enblad, P. (2009). Intracerebral monitoring in comatose patients treated with hypothermia after a cardiac arrest. Acta Anaesthesiologica Scandinavica, 53(3), 289-298. doi: 10.1111/j.1399-6576.2008.01885.x Peberdy, M. A., Callaway, C. W., Neumar, R. W., Geocadin, R. G., Zimmerman, J. L., Donnino, M., . . . Kronick, S. L. (2010). Part 9: Post-cardiac arrest care: 2010 american heart association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation, 122(18 Suppl 3), S768-86. doi:10.1161/CIRCULATIONAHA.110.971002 Pfeifer, R., Jung, C., Purle, S., Lauten, A., Yilmaz, A., Surber, R., . . . Figulla, H. R. (2011). Survival does not improve when therapeutic hypothermia is added to post-cardiac arrest care. Resuscitation, 82(9), 1168-1173. doi: 10.1016/j.resuscitation.2011.05.024Prior, J., Lawhon-Triano, M., Fedor, D., Vanston, V. J., Getts, R., & Smego, R. A.,Jr. (2010). Community-based application of mild therapeutic hypothermia for survivors of cardiac arrest. Southern Medical Journal, 103(4), 295-300. doi: 10.1097/SMJ.0b013e3181d3cedb Reinikainen, M., Oksanen, T., Leppanen, P., Torppa, T., Niskanen, M., Kurola, J., & Finnish Intensive Care Consortium. (2012). Mortality in out-of-hospital cardiac arrest patients has decreased in the era of therapeutic hypothermia. Acta Anaesthesiologica Scandinavica, 56(1), 110-115. doi: 10.1111/j.1399-6576.2011.02543.x; 10.1111/j.1399-6576.2011.02543.x Roger, V. L., Go, A. S., Lloyd-Jones, D. M., Adams, R. J., Berry, J. D., Brown, T. M., . . . American Heart Association Statistics Committee and Stroke Statistics Subcommittee. (2011). Heart disease and stroke statistics--2011 update: A report from the american heart association. Circulation, 123(4), e18-e209. doi:10.1161/CIR.0b013e3182009701 Storm, C., Nee, J., Roser, M., Jorres, A., & Hasper, D. (2012). Mild hypothermia treatment in patients resuscitated from non-shockable cardiac arrest. Emergency Medicine Journal : EMJ, 29(2), 100-103. doi: 10.1136/emj.2010.105171 Stub, D., Bernard, S., Duffy, S. J., & Kaye, D. M. (2011). Post cardiac arrest syndrome: A review of therapeutic strategies. Circulation, 123(13), 1428-1435. doi:10.1161/CIRCULATIONAHA.110.988725 Stub, D., Hengel, C., Chan, W., Jackson, D., Sanders, K., Dart, A. M., . . . Kaye, D. M. (2011). Usefulness of cooling and coronary catheterization to improve survival in out-of-hospital cardiac arrest. The American Journal of Cardiology, 107(4), 522-527. doi: 10.1016/j.amjcard.2010.10.011 Testori, C., Sterz, F., Behringer, W., Haugk, M., Uray, T., Zeiner, A., . . . Losert, H. (2011). Mild therapeutic hypothermia is associated with favourable outcome in patients after cardiac arrest with non-shockable rhythms. Resuscitation, 82(9), 1162-1167. doi: 10.1016/j.resuscitation.2011.05.022 Tiainen, M., Poutiainen, E., Kovala, T., Takkunen, O., Happola, O., & Roine, R. O. (2007). Cognitive and neurophysiological outcome of cardiac arrest survivors treated with therapeutic hypothermia. Stroke; a Journal of Cerebral Circulation, 38(8), 2303-2308. doi: 10.1161/STROKEAHA.107.483867 van der Wal, G., Brinkman, S., Bisschops, L. L., Hoedemaekers, C. W., van der Hoeven, J. G., de Lange, D. W., . . . Pickkers, P. (2011). Influence of mild therapeutic hypothermia after cardiac arrest on hospital mortality. Critical Care Medicine, 39(1), 84-88. doi: 10.1097/CCM.0b013e3181fd6aef Summary of Evidences1stAuthorStudy design /Control/LOESampleTH vs NTType of CA rhythmSite of CACooling deviceInduction time from ROSCTarget temp in TH group /Mean tempCooling durationMean temp in NT group Survival OutcomeTH vs NTNeurological OutcomeTH vs NTTiainen (2007)Randomized prospective trial(I B)N = 70TH = 36NT= 34Shockable*OHCAExternal deviceNA32-34C/33+1C24hrsMean NA^NTMean NA[SO] ( 28 vs 22, p = 0.226)93% vs 78%, no P valueCognitive outcome:Intact or subtle deficit67% vs 44%, NSKim (2007)Randomized controlled trial(I C)N = 97TH = 49NT = 48All rhythms*OHCASC and CSNA32-34C/Mean NANA×NTMean NA[SO] VF group: 66% vs 45%, p= NSNon VF group:6% vs 20%, p = NSAdjusted OR for survival = 0.91, 95% CI=0.28 to 2.96Awakening in VF patients:69% vs 45%, P = 0.15Awakening in Non-VF patients:9% vs 23%, P = 0.13Granja (2011)Before and after(II B)N = 130TH = 55NT =75All rhythmsIN and **OHCASC and CS4+2.25hrs32-34C/Mean NA15.1+4.1hrs×NTMean NA[SO] 60% vs 39% ,P= 0.1626 vs 21 patients, NSReinikainen 2012Retrospective observational with HC (III B)N = 3958TH = 3072NT = 886ShockableOHCANANANA32-34C/Mean NANA×NTMean NA[M]51.1% vs 57.9%, P< 0.001Adjusted OR = 0.54, 95 % CI= 0.45-0.64, P < 0.001)NAPrior(2010)Retrospective cohort with HC(III B-/C)N = 456TH = 44NT = 368All rhythmsIN and **OHCASC2.8hrs (0.2-7.832-34C/Mean NA9-28hrs×NTMean NA[SO] Within TH group:Shockable vs non shockable61% vs 24%, P < 0.0543% vs 13%, P < 0.001Pfeifer(2011)Retrospective observational with HC(III B)N = 210TH = 143NT = 67All rhythmsIN and **OHCASC and CS4-6hrs32-34C/33+1C24hrsMean NA^NTMean NA[SO] All patients:48.2% vs 44.8%, P = 0.659For shockable :26.4% vs 28.6%, P = 0.807For non-shockable:70.4% vs 56.4%, P = 0.149)Better in VF patient within TH group (p < 0.001)Testori,(2011)Retrospective cohort with HC(III B)N = 374TH =135NT= 239Non Shockable*OHCASC and CS1.4hrs32-34C/33+1C24hrsMean NA^^NTMean NA[SO] Adjusted OR = 0.56, 95% CI, 0.34 – 0.93Adjusted OR = 1.84, 95% CI, 1.08 – 3.13Storm (2012)Prospective observationalwith HC(III B)N = 175TH = 87NT = 88Non ShockableIN and **OHCASC and CSNA32-34C/Mean NA24hrsMean NA×NTMean NA[SO] Adjusted HR 0.98, 95% CI = 0.53-1.5, p = 0.6327.5% vs 18.2%, P = 0.175Stub (2011)Retrospective observationalwith HC(III B)N = 125TH = 81NT = 44Shockable**OHCASC and CSNA32-34C/Mean NANA×NTMean NA[SO] 64% vs 39%, p <0.01Unadjusted Odds ratio 2.7, 95% CI = 1.1 – 6.4 , P = 0.0257% vs 29%, p < 0.01Van der (2011)Retrospective observationalwith HC(III B)N = 5317TH= 3770NT = 1547All rhythms**OHCANANA33-36.4CNA^^NTMean NA[M] 65% vs 72%, p = NA Adjust OR= 0.8, 95% CI= 0.65 – 0.98, p = 0.29NALOE= Level of evidence; CA=Cardiac arrest; ROSC=Return of spontaneous circulation; SO=Survival Outcome; M=Mortality Outcome; All rhythms= VT/VF/PEA/Asystole; TH=Therapeutic hypothermia; Shockable =VT/VF; Non Shockable=PEA/Asystole; NT=Normothermia; ^ hyperthermia controlled; ^^ hyperthermia not controlled; × hyperthermia information NA; *OHCA=Out of hospital cardiac arrest(witnessed); **OHCA=Out of hospital cardiac arrest(witnessed and un witnessed);IN=In-hospital; NS=Not significant; NA=Not available/assessed; SC=Surface cooling; CS=cold saline/fluids; HR=hazard ratio; HC=HistoriJohns Hopkins Nursing Evidence-Based Nursing PracticeIndividual Evidence Summary ToolEBP Question: In adult post cardiac arrest patients, who have return of spontaneous circulation, does therapeutic hypothermia compared with normothermia, improve mortality and neurological outcomes.Key terms: Therapeutic hypothermia (TH), Normo-thermia(NT),Return of spontaneous circulation (ROSC), Cardiac arrest (CA), Intracranial hemorrhage (ICH), Out of hospital cardiac arrest (OHCA), Systolic blood pressure (SBP), Ventricular fibrillation (VF), Ventricular tachycardia (VT), Glasgow coma scale (GCS), Mini mental state examination (MMSE), Cerebral perfusion category (CPC/CPC), Systolic blood pressure (SBP), Simplified acute physiology score (SAPS). Not available (NA), Not significant (NS), Hemodynamic (HD), cerebrovascular accident (CVA), Basic life support (BLS), Diabetes Mellitus (DM), Chronic Obstructive pulmonary disease (COPD), Acute physiology and chronic health evaluation (APACHE)Definition: ROSC (return of spontaneous pulse that is sustained for at least 20 minutes), No flow time (Time from collapse to start of BLS), low flow time (Time from start of life support until restoration of ROSC).#Author & DateStudy DesignAnd Key elements Sample, Sample Size, & SettingStudy findings that help answer the EBP questionTH versus NT groupLimitationsLOE & Quality1Tiainen (2007)Design: Randomized prospective studyStudy Duration: March 1997 and June 2000Control Group: Normo-thermia<38 CIntervention Details Target temperature: 33 + 1CInduction phase: NA TH initiation site : Hospital TH phase: 24 hrs Rewarming phase: Passive, at rate of 0.5 C per hour over 12 hours.Cooling device : External device (TheraKool Kinetic Concept)Variables controlled: Both groups received standard ICU protocol treatment. Sedation and shivering/seizure prevention for 32 hoursMAP > 80 mmHgNormoglycemia < 10mmol/LHOB elevation > 30 Neuropsychologist was blinded Procedural and monitoring controls.Both groups are statistically comparable for age, gender, education, BLS, ACLS time, GCS, blood glucose, type of initial rhythm, acute MI, ischemia, neuron-specific enolase, temperature on admission and coronary interventions.More than one test for neuro evaluation to prevent monooperation bias.Instruments used for outcome measurementTemperature monitoring technique: Urinary bladder prehensive neurophysiological examination.Pittsburgh CPC Cognitive assessment Wechsler Adult intelligence scale-revised (WAIS-R)Learning and memory domain assessment: Wechsler memory scale revised (WMS-R) and auditory verbal learning test (AVLT). Executive functioning assessment: Modified stroop test, trial making test, verbal fluency task, timed calculation taskQuantitative electroencephalogram (Q-EEG) Auditory P300Statistical significance : 0.05 (one or 2 tail not mentioned)Setting : Helsinki university hospital Inclusion criteria Age: 18-75 Initial rhythm: VF/Non- perfusing VTCardiac origin of arrestSite of CA: OHCA (witnessed)No flow time: 5-15minsLow flow time <60mins ROSC definition: No timeline defined.GCS < 9Exclusion CriteriaPregnancyTemp < 30 C O2 sat < 85% for > 15 minsMAP < 60 for > 30 mins Response to verbal commands after ROSC.Terminal illness leading to arrest.CoagulopathyFactors that makes follow-up unlikely Sample Size N = 70TH group= 36Control group= 34Randomization: Yes, technique not mentionedPower analysisNASurvival Outcomes Survival to discharge, at 3 and 6 months (n= 50, 28 vs 22, P = 0.226)Neurophysiological Outcomes at 3 months Neurophysiological outcomes at 3 months in 45 of 50 survivors (27 versus 18). Good neurological outcome was observed in (93% versus 78%, no P value)Cognitive outcome:Severe deficit (15% versus 28%, NS)Moderate deficit (19% verse 28%, NS)Intact or subtle deficit (67% versus 44%, NS)Learning and memory domain: NSExecutive functioning deficit: NSQ-EEG NS.P 300 was significantly higher in TH.P300 correlated with CPC class (r = -0.37, P <0.016)Significant delay in publication.Sampling issuesThere is no evidence if the groups were similar in terms of their acuity level at the time of CA that could mask the improvement.Single center study. Patient population of the hospital may not be representative of entire CA population.Treatment issues Mean and SD values for TH induction time, time to target temperature and mean temperature in control group not provided. Thus, consistent application of TH is difficult to ascertain.No evidence if controls were applied beyond 32 hrs of CA. Neurological sequel can be observed within 72 hrs from CA.Instrument and measurement None Analysis and resultsRandomization process detail not mentionedNeurophysiological and cognitive outcomes were favorable in TH group but could not reach statistical significance may be because of sample size, sample mortality at 30 days, and further sample attrition for specific neurological exam at 6 months. Power analysis not done to support sample size and minimize threat of type II error.Treatment induction time and time to target temperature from CA is not provided. Thus, it is difficult to comment if delay in initiating or reaching optimal target temperature were contributor to NS results.P value for neurological outcomes is not provided for statistical conclusion.In both groups there were few patients with ischemic stroke that could have masked the neurological recovery from CA.Generalizability is limited due to: Probably inadequate sample size. It only included patients with VF (witnessed) arrest. Results cannot be generalized to CA patients with non-shock able rhythm.Stringent exclusion criteria to control potential confounders has improved the study validity but has limited generalizability even within VF group who do not share same clinical characteristics.Absolute recommendations related to induction time cannot be drawn.I B2Kim(2007)Design: Randomized controlled trial Study Duration: Nov 2004 to February 2006Control Group : Normo-thermiaIntervention Details: Standard care plus hypothermia. Target temperature: 32-34CInduction phase: Soon after resuscitation TH initiation site : Out of hospital TH phase: Continuation of pre hospital cooling was left at admitting physician discretion during hospitalizationRewarming phase: Not providedCooling device: 2 L of NS at 4 C at pre hospital and surface cooling in hospital. Variables controlled: Groups comparable for age, gender, witnessed CA, CPR before paramedic’s arrival, HR, systolic, blood gases, O2 saturation, BP, and pressors at arrival and 12 hours of hospitalization.Other controls during intervention for both groupSedation and seizures prevention MAP > 80 mmHgNormoglycemia < 10mmol/LHOB elevation > 30 Instruments used for outcome measurementTemperature monitoring technique: Esophageal temperature sensor in field and/or tympanic temperature probe in hospital temperature Statistical significance : 0.05Setting : Seattle, involving 7 paramedic units and 7 acute care hospitals.Inclusion criteria Age: ≥ 18 yearsInitial rhythm: All rhythms Non traumatic CA (witnessed)Site of CA: Out of hospital Exclusion CriteriaTraumatic cardiac arrestFollowing commands post ROSCTemperature < 34 CRe-arrest during treatment Sample Size Enrolled N = 125TH group= 63Control group= 62Randomization: Yes, over phone while patient still in field, Balanced block of 4Survived up to admissionN = 97TH group= 49Control group= 48Power analysisNone Survival Outcome at dischargeVF group: 66% versus 45% (P= NS)Non VF group: 6% versus 20% (P = NS)Field those who received field cooing alone OR for survival to discharge = 1.92 (95% CI= 0.46 to 8.0) adjusted for hospital cooling and interaction terms. For those received hospital cooling alone OR for survival to discharge = 0.91 (95% CI=0.28 to 2.96) adjusted for field cooling and interaction terms.Neurophysiological Outcomes at hospital Awakening in VF patients: 69% versus 45%, P = 0.15Awakening in Non-VF patients: 9% versus 23%, P = 0.13Sampling Issues 65 eligible patients were not randomized and enrolled in the study. And 23 cases were not considered for screened for eligibility. It is plausible that these patients differ from those enrolled in the study.28 patients enrolled in the study could not survive up to admission. Treatment issues.Treatment controls were poor for RCT.End point for temperature not defined before study.Highly inconsistent administration of treatment. 8 subjects did not receive pre-hospital cooling at all, and only 12 received full 2L of cold fluid. Once at hospital, patients were treated as per hospital protocol. There was some crossover of subjects from control to TH group and vice versa. 60 of 97 patients received hospital cooling regardless of field cooling.All hospitals had variable protocols. Duration of hospital cooling, induction timing, mean temperature during hospital, rewarming details not provided to ensure consistency.Random and inconsistent use of pancuronium and midazolam for shivering control in TH group for shivering control.Instrument and measurement Inconsistent instrument used for temperature monitoring in field and in hospitalNo valid instrument used for neuro outcomes, instead “awakening” was used as a parameter, and that too was assessed through charts. Analysis and resultsThe study mainly focuses on outcomes of pre hospital cooling. Discharge and neurological evaluation was on sub sample.No power analysis performed. Sample size not adequate to render conclusion on effect of TH on survival of neurological outcome. No indication if pyrexia was prevented in controlled group that could have attributed to poor outcome. And mean SBP, blood gases, O2 saturation, were not compared for the entire duration of treatment. These variables can be potential confounders.Several missing entries of temperature in the record. Researcher does not specify how missed data was handled.Results of awakening in non-shockable group and adjusted OR for hospital cooling is inconsistent with all other studies.GeneralizabilityLimited because of serious threats to internal and statistical validity.1 C3Granja 2011Design: Before and after Study Duration: 2004-2009 (Before, Oct 2004-Mar 2006) and (After, April 2006-Sep 2009)Control Group : Normo-thermiaIntervention Details Target temperature: 32-34 CInduction phase: 4hrs (SD + 2.25)TH initiation site : ICUTH phase: Target 12-24hrs, mean 15 hrs (SD + 4.11hrs)Rewarming phase: Passively, for 5hrs (SD + 2.33 hrs)Cooling device : Cold normal saline infusion (4C at 30ml/hr) and surface cooling Variables controlled: Shivering control measures in TH groupGroups were comparable for age, gender, reason for admission, no flow time, type of initial rhythm.Instruments used for outcome measurementPittsburgh CPC scale for neurological evaluation. Esophageal temperature probe Statistical significance : 0.05 Setting : General ICU at urban general hospital Inclusion criteria Age: >18Initial rhythm: All rhythms Origin of arrest: AllSite of CA: In and out of hospital (witnessed and non-witnessed)Exclusion CriteriaEsophageal temperature less than <32Traumatic brain injuryStatus epilepticsPregnancy Refractory hypotension SBP < 80Bleeding diathesis Incomplete data in fileConscious after CA.Sample Size N = 130TH group= 55Control group= 75Randomization: NonePower analysisNot done Survival to discharge Survival outcome in TH group is better but NS60% versus 39% (P = 0.16)Neurophysiological Outcomes after 6 months.CPC evaluated only on 56 patients 28 in each group.26 versus 21 patients presented favorable neurological outcomes (P = NS)Sampling issues Non randomized sampling can lead to selection bias.Treatment issues Study does not inform if control and TH groups were handled in a similar way; if pyrexia was restricted in control group; and if MAP, glycemic control, oxygenation parameters were same in both. Treatment is elaborated but quality control measures for standardization of intervention is not mentioned in the article.Instrument and measurement issues Mean temperature attained in intervention and control group is not provided.Person conducting neurological evaluation was not blinded that can potentially add bias.Analysis and resultsPower analysis is not conducted to support adequacy of sample size. NS findings might be due to type II error secondary to small sample.As per mean duration of TH treatment 50% of patient were cooled for less than 15 hrs. short duration might have contributed to NS results Sample attrition at 6 months for neurological leaves a very small sample size. Thus, results are prone to type II error.Date was collected retrospectively and monitoring controls cannot be assured. Not all potential confounders or effect modifiers listed above were adequately controlled.Generalizability Is limited because of threat to statistical validity due to low sample size and lack of control on potential confounders.II B4Reinikainen (2012)Design: Retrospective observational studyStudy Duration: 2000-2008, (pre hypothermia era, 2000-2002) and hypothermia era (2003-2008)Comparison group : Normothermia Historical controlIntervention Details Target temperature: 32-34C Induction phase: NATH initiation site : NATH phase: 24 hours mostlyRewarming phase: NACooling device : NAVariables controlled: NAInstruments used for outcome measurementNA Statistical significance : Not mentioned Setting : Finnish intensive care settings (20 hospitals).Inclusion criteria Age: 18+Initial rhythm: VF or non perfusing VT.Cardiac origin of arrest: Site of CA: OHCA No flow time 5-15 minuets Low flow time < 60 minutesExclusion CriteriaIn-hospital CASample Size N = 3958TH group= 3072Non TH group= 886Randomization: None Power analysisNAHospital mortality51.1% versus 57.9% (P < 0.001)Hypothermia was associated with a decreased risk mortalityOR = 0.54 (95 % CI= 0.45-0.64, P < 0.001) adjusted for the SAPS II score and gender. This effect was consistent for both young and older patients... Neurophysiological Outcomes Not assessed Sampling issuesHuge variation in number of patients in the groups (3702 versus 886).Selected population of CA with shockable rhythms is included that is not representative of all CA rmation not available on peri arrest factors like no flow, low flow time, witnessed and non-witnessed status of CA. It cannot be commented if groups were comparable at baseline.Treatment issues Mean temperature in groups, induction time, optimal temperature, duration of TH and rewarming phase details not provided in the study. It is difficult to determine how well treatment was executed.Instrument and measurement issues Instrument used for temperature monitoring is not mentioned. Core versus shell temperature monitoring results are different.Analysis and results issues Level of significance not mentioned. Tight glycemic control practices were also widely adopted during TH period that could have confounded the true association between TH and outcomes which is not controlled in the analysis. New resuscitation guidelines were published in 2003 that might have intensified resuscitation practice and post ROSC care in general during interventions period of the study, attributing to positive results. This is not acknowledged in the study. Mean temperature in control group is also not provided. And there is no indication if pyrexia in pre TH group was controlled that could have adversely effected the outcomes in that group.Power analysis not providedGeneralizability Findings can be generalized with caution to CA patient with OHCA with shockable rhythm only In absence of treatment related details it is difficult to extrapolate exact recommendations about treatment details like induction time, optimal temperature, duration of TH and rewarming phase. Thus, translation of research into practice seems obstructed.III B5Prior(2010)Design: Retrospective cohort study with historic controls Study Duration: 2002-2004Comparison Group : Normo-thermia Historical controlIntervention Details Target temperature: 32 C to 34 CInduction phase: Target < 4 hrsMean 2.8 hrs (range 0.2 -7.8) from ROSC Target achieved in mean 7.2 hrs (range 0.8 to 15.1 hrs)TH initiation site : In hospital TH phase: 24hrs (range 9 to 28 hours)Rewarming phase: Passive Cooling device : surface cooling Ice and cooling blankets Variables controlled: Shivering measures for TH groupGlycemic control in TH group Hyperthermia controlled after rewarming in TH group onlyInstruments used for outcome measurementBladder temperature technique Pittsburgh CPC scale for neurological evaluation Statistical significance : 0.05Setting : 3 community hospitals in Scranton, Pennsylvania Inclusion criteria Age: >18 years Initial rhythm: All rhythms Cardiac origin of arrest: Site of CA: Out of hospital and in hospital (witnessed and non-witnessed)GCS: Coma Exclusion CriteriaPregnancycardiogenic shockSBP of less than 90 mm Hg despite epinephrine infusion)Non CA cause of coma Sample Size N = 456TH group = 44Control group = 368Randomization: None Power analysisNASurvival to dischargeWithin TH group: Survival of patients in VF/VT group is higher than in non-shockable rhythm 61% versus 24%, P = < 0.05Good Neurophysiological Outcomes at discharge 43% versus 13% (P < 0.001)Patients with a good CPC score were less likely to achieve goal temperature within 8 hours (P <0.05) and were cool longer than 24 hrs (P < 0.02)Sampling Assignment to groups was not random that pose selection bias.Sample size in TH group is very less compare to control.Treatment issuesHuge variation in induction and TH maintenance time as range is wide. Treatment applied inconsistently as 4.4% of temperature readings were above 34 C and 16.4% below 32C. Comparative analysis not provided for TH and control groups for key characteristics like demographic, acuity level, no flow and low flow time, duration of resuscitation, type of initial rhythm, MAP, O2 Saturations.TH and controls were not treated similarly Inclusion and exclusion criteria were not applied to controls. Thus groups might not be homogenousSeizure, pyrexia and glycemic control not applied to controlsInstrument and measurement Definition of good neurological outcome was set different for controls versus TH group. CPC was not used for neuro evaluation in controls group. CPC score was assigned through chart review and the assessor was not blinded.Analysis and resultsPyrexia and glycemic control was not observed in control group that could be the explanation of poor outcome in this group.General post CA ROSC care also improved under new guidelines during intervention period. Favorable outcomes might be attributable to that.Power analysis is not done and groups are not equal.Survival to discharge analysis does not include comparison of TH with control group. It only includes comparison of outcomes among VT/VF and non shockable patients within TH parison of neurological outcomes in VT/VF versus non shockable rhythms is not provided. It is difficult to conclude if all benefited equally. Groups were not homogenous so results in favor of TH might be due to specific characteristics in control group patients.Generalizability Lack of homogeneity, decrease sample size in TH group, inconsistent treatment of TH and control group, different operational definition of neurological outcome for TH versus control group, threatens the study validity. Results should be generalized with extreme caution.III B6Pfeifer(2011)Design: Retrospective studyStudy Duration: 2003-2010Comparison Group: Normo-thermia(< 38C ) Historical controlIntervention Details Target temperature: 33 + 0.5 CInduction phase:4-6 hrs of CATH initiation site : In hospital TH phase: 24 hrs Rewarming phase: Passive and active (0.3 C per hour)Cooling device : Surface or intravascular cooling methodVariables controlled: Shivering control measures in THPyrexia control in bothStandard ICU care in both Instruments used for outcome measurementBladder temperature probePittsburgh CPC Statistical significance : O.05Setting : Single center registry Inclusion criteria Age: 18+Initial rhythm: All rhythms Non trauma cardiac arrest.Site of CA: In and out of hospital (witnessed and non-witnessed)Low flow time: < 60 minsHypoxia time < 15 mins in non-witnessed CA.Exclusion CriteriaRegained consciousness within 1hr of CA.TH induction time more than 6 hrs from ROSCRefractory hemodynamic instability without vasopressors.Sample Size N = 210TH group= 143Control group= 67Randomization: NonePower analysisNAMortality rate at 30 days from CATH group did not have any mortality benefit even if they were comparatively young and less severely illAll patients:48.2% versus 44.8% (P = 0.659)For VF patients : 26.4% versus 28.6% (P = 0.807)For non-shockable: 70.4% versus 56.4% (P = 0.149)Statistically significant survival outcomes was observed in only those who had hypoxia time < 6 minutes, low flow time < 20 mins and VF as a primary rhythm. (p = 0.044)Comparison of patients with good and poor CPC score within TH group Patients with good neurological outcome in TH group were young (58.6 versus 64.6, P= 0.014), more often had VF as primary rhythm (P < 0.001), had lower SAP II score (P = 0.04), had shorter duration of CPR (P < 0.001) and were cooled for longer period (P = 0.003) after adjusted for SAP, no flow/low flow time, duration of CPR.Neurological outcome in comparison with control group NA Sampling There are several factors potentiating selection bias.Non randomized assignment of individuals to group. On duty doctor decided for TH or control based on inclusion that could have led to selection bias.Only those patients who survived 48 hours after CA in the groups were apprised in the analysis. Those who survived and included in analysis might be less critical and different in particular characteristics then those who were not.HT group has more patients than controlThe groups were not comparable. Patients in control group were significantly older, and had higher acuity level (high SAP score) and had more patients with in-hospital CA that goes in favor of TH group.There is no information if groups were comparable in no flow and low flow time.Treatment issuesMean temperature in control group is not available to verify if claim of pyrexia control was actually achieved in all patients.Instrument and measurement issues CPC scores were assigned through documents review, telephone interview with patients and discharge letters, instead of clinical examination. Further, assessor was not blinded that could have added bias to neurological results.Analysis and resultsPower analysis is not performed. Inadequate power might have led to NS findings due to risk of type II error. There is no evidence if potential confounder like hyperglycemia, MAP, O2 saturation were controlled through exclusion criteria, group comparison or regression analysis.Neurological evaluation is presented for TH group only, comparison with control group is not provided.Generalizability is limited due to:Statistical validity is questioned as power analysis is not performed and findings may be prone to type II error. Single center study.III B7Testori(2011)Design: Retrospective cohort study registry based Study Duration: 1992-2009Comparison Group : Normothermia Historical controlIntervention Details Target temperature: 33 C + 1 CInduction phase: 1.4 hrs (IQR 0.71 to 2.33) from ROSCTarget achieved in 2.9 hrs (IQR 1.48 to 4.11)TH initiation site : TH phase: 24 hrs Rewarming phase: NACooling device : Surface, invasive or combined cooling techniques Variables controlled All TH patients received standard sedation and muscle relaxant to prevent shivering.Instruments used for outcome measurementInfrared tympanic thermometer, esophageal probe, or foleys catheter Pittsburgh CPS for neurological evaluation Statistical significance : 0.05SettingSingle tertiary care setting Inclusion criteria Age: 18+ Initial rhythm in CA: Non-shockable Cardiac origin of arrest: Non traumatic Site of CA: Out of hospital (witnessed)Exclusion CriteriaGCS > 8CPS > 2 before CATemp <30 CCVA associated CASample Size N = 374TH group= 135Control group= 239Randomization: NonePower analysisNASurvival at 6 months Mortality rate was significantly lower in TH groupOR = 0.56 (95% CI, 0.34 – 0.93)adjusted for age, gender, cause of CA, bystander BLS (Y/N), no flow time, low flow time, history of DM, COPD, epinephrine doses, GCS on admission Neurophysiological Outcomes at six months Neurological outcomes were better in TH group OR = 1.84 (95% CI, 1.08 – 3.13) adjusted for potential confounder listed above.Sampling issues Several factors might have caused selection bias.667 eligible patients with CA who died in initial 24 hours were excluded during registry review and only survivors (374patients) beyond 24 hrs were included in analysis. This might have excluded pool of most critical patients from the study. There is no indication in the study if those who were not included in the study differed in particular characteristics from those included.Non randomized assignment to TH and normo-thermiagroup.Cooling decision was at discretion of on call doctor, who might not have cooled patients with expected bad outcomes.374 patients for the period of 1992-2009 is a very small sample Groups were unequal. Above factors are threats to internal validity.Treatment related issues No quality control measures to ensure treatment standardization over the period.Instrument and measurement issues 3 separate temperature monitoring devices were used. Different cooling devices were used Collection of data from registry thus had missing data. There is no specification how CPC evaluation was performed.CPC validity is not tested for retrospective application.Analysis and resultsPower analysis is not doneBecause of retrospective observational study, several confounders like MAP, oxygenation status, patient’s acuity level, glucose level, hyperthermia, baseline GCS post CA were not controlled.Groups were not treated equally. Patients in comparable group had a trend towards hyperthermia which may have contributed to the poor outcome in the group that threatens statistical validity.There is no specification how missed temperature recordings were handled. Generalizability Findings should be generalized with caution because:Exclusion of post CA fatalities within 24 hours limits the sample to relatively stable patients who may not be true representative of actual population of CA with initial non-shockable rhythm.Out of hospital, witnessed cardiac arrest patients were enrolled that limits generalizability to in hospital and non-witnessed CA patients who may be more critical. Single center study.III B8Storm(2012)Design: Prospective observational study.Study Duration: 2002-2010Comparison Group : Normo-thermia Historical controlIntervention Details Target temperature: 33 CInduction phase: Time from ROSC: NATarget achieved in 3 to 5.33hrs TH initiation site : Hospital TH phase: 24 hrs Rewarming phase: Controlled at 0.25 C per hour.Cooling device: Surface and invasive cooling devices.Variables controlled All TH patients received standard sedation and muscle relaxant to prevent shivering.All TH patients received standard post ROSC care.Instruments used for outcome measurementPittsburgh CPS for neurological evaluation Statistical significance : 0.05Setting: University hospital Charite BerlinInclusion criteria Age: 18+ Initial rhythm in CA: Non-shockable Cardiac origin of arrest: Non traumatic Site of CA: Out and In hospital Exclusion CriteriaNAOver all Sample Size N = 387TH group= 201Control group= 186Randomization: NAStudy presents sub analysis of patients with non-shockable rhythmN = 175TH group= 87Control group= 88Power analysisNASurvival at discharge Cox regression showed NS improvement in TH groupHazard ratio 0.98, 95% CI = 0.53-1.5, p = 0.63 adjusted for APACHIE score, low flow time, ICU stay.90 days Kaplan Meier analysis revealed NS results (log rank test P = 0.82Neurophysiological Outcomes at discharge TH was not associated with significantly improved neurological outcomes in TH group27.5% versus 18.2% (P = 0.175)TH group had decreased low flow time. Sampling issues 175 patients for the duration of 8 years seems to be small. The study does not specify how patients were excluded.Treatment related issues The treatment execution might have differed over the prolong study period.Instrument and measurement issues There is no information how temperature was monitored Temperature monitoring devices used is not specified thus instrument validity cannot be ensured.Analysis and resultsMean temperature in control group is not provided.Power analysis not conducted. NS result may be due to inadequate sample and type II error. Details not provided if control group also received standard post ROSC treatment (excluding hypothermia) as of TH group. Thus, several confounders like MAP, oxygenation status, patient’s acuity level, glucose level, and hyperthermia could have confounded the results that are not controlled during regression analysis. Neurological evaluation is provided at the time of discharge only. Thus, improvement at 90 days or 6 months was not captured in the results.Generalizability May be limited because of Single center study Probably inadequately powered study. III B9Stub(2011)Design: Retrospective observational study Study Duration: 2002-2009Comparison Group : Normo-thermia Historical controlIntervention Details Target temperature: 32-34CInduction phase: From ROSC NATarget achieved in NATH initiation site : NATH phase: NARewarming phase: NACooling device: Surface and intravenous cooling interventions.Variables controlled Standard ROSC care for both groups Instruments used for outcome measurementPittsburgh CPC NAStatistical significance : 0.05SettingAlfred hospital Inclusion criteria Age: 18+ Initial rhythm in CA: VF/VTOnly cardiac origin of arrest: Site of CA: Out of hospital (witnessed and non-witnessed)Exclusion CriteriaGCS All non-cardiac causes of CA (stroke, drug over dose, trauma)Sample Size N = 125TH group= 81Control group= 44Randomization: NonePower analysisNASurvival at discharge64% versus 39% (p <0.01)Survival rate was significantly better in TH group Unadjusted Odds ratio 2.7, 95% CI = 1.1 – 6.4 , P = 0.02 Neurophysiological Outcomes at discharge 57% versus 29% (p < 0.01)Sampling issues During intervention period, the institution simultaneously introduced coronary artery disease evaluation and the need for coronary intervention as standard post CA ROSC care along with TH. The study is prone to selection bias because it limits participant who had CA of cardiac origin as they wanted to examine impact of 2nd intervention on survival outcome as well.Non randomization of patients to TH versus control groupTreatment related issues Intervention details like induction time, TH duration, mean temperature in TH and control group, rewarming strategies, cooling devices used etc. is not provided in the article.Instrument and measurement Hospital records were used for data collection.CPC scores were assigned through chart review instead of clinical evaluation.CPC validity is not tested for retrospective application.Analysis and results must be interpreted with caution because of serious threats to internal validity. Effect of coronary intervention was not adjusted during statistical analysis and only univariate analysis is presented.Significantly high number of patients in TH group also received coronary interventions like angiography (p < 0.01) and PTCA (p = 0.03). Keeping in view this potential confounder and in absence of adjusted analysis, TH cannot be labeled as sole predictor of positive outcome. New resuscitation practices were also adopted during treatment period like compression to ventilation ration and uninterrupted compression. Moreover, duration of resuscitation was short in treatment group that could perhaps have attributed to favorable results in TH group.CPC scores to ascertain neurological outcomes were assigned through chart review. Thus, precision of data is doubtful. Power analysis not conducted.Generalizability Can be limited because of following threats to internal validity due to:OHCA patients, with VF/VT as initial rhythm and with cardiac origin of CA were included who may not be representative of all CA patients with VF/VT as an initial rhythm Inadequate information about the treatment seriously compromises the recommendations.Single center studyIII B 10Van der (2011)Design: Retrospective multicenter observational study.Study Duration: 1999-2009 Historical controlBefore Implementation Group : Minimum temperature 35.5 C (IQR 34.3 C – 36.5 C)Maximum temperature 37.8 C (IQR 36.8 C – 38.5 C)After implementation group Temperature: Minimum temperature 33 C (IQR 32 C – 35.4 C)Maximum temperature 36.4 C (IQR 35.1 C –37.6 C)Induction phase:Time from ROSC: Target achievedTH initiation site : TH phase: 12 – 24 hrsRewarming phase: NACooling device : NAVariables controlled NoneInstruments used for outcome measurementNAStatistical significance : 0.05Setting: 59 ICUs of Netherland Inclusion criteria Age: 18+Initial rhythm in CA: Both shockable and non shockable Site of CA: In and Out of hospital Exclusion CriteriaCA in ICU GCS > 8 post ROSCCA patients in initial 3 months of implementation period.Cardiothoracic and burn patients.Over all Sample Size N = 5317TH group = 3770Control group= 1547Randomization: NAPower analysisNASurvival outcomes at discharge Mortality of CA patients before and after treatment period 65% versus 72%, p = NAOdd ration for hospital mortality for CA patients after mild therapeutic hypothermia is OR= 0.8, 95% CI= 0.65 – 0.98, p = 0.29 adjusted for SAP score, age, gender, in and out of hospital CA, and propensity score. Propensity score is the probability of a patient to be part of before or after implementation group.Patients with in hospital CA has significantly high mortality compare to OHCA OR = 1.16, p = 0.26Neurophysiological Outcomes at discharge NASampling issues Data was extracted from registryNon random selection of ICUs. Treatment related issues All institutions had their own TH protocol therefore treatment standardization could not be maintained. Thus, there was a wide variation range of minimum and maximum temperature and IQR in TH group. Moreover, maximum temperature in the group is almost close to normothermia.Mean induction time, optimal temperature time, duration of TH, cooling devices used in not mentioned in the article.Instrument and measurementInstrument used for temperature monitoring is not specified. Consistency in findings cannot be assured.Result and analysisDue to wide variation in temperature end point in TH group, it cannot be ascertain whether hypothermia or lower threshold normo-thermiawith controlled pyrexia was associated with positive outcomes 41% of patients in pre implementation group had experienced pyrexia during first 24 hrs that could have contributed to poor outcome in control group. In post implementation group, practices like pyrexia prevention, glycemic control, coronary interventions and general post ROSC care had improved because of new guidelines. These covariates could have contributed to improved outcomes but were not controlled in the study.Researcher could not get information on, mean no flow and low flow timings. Systematic difference in these variables among group might have attributed to the results.The researcher does not inform the neurological outcomes of patients Generalizability is restricted Inadequate information about treatment details and wide variation in target temperature in TH group restrict any absolute recommendation for future.Findings cannot be generalized to in- patients CA setting because mortality outcomes are not significant in that group. Survival outcomes are not provided as per type of initial CA rhythm. Thus, we are not sure if the treatment is effective in patients with both shockable and non-shockable rhythm. III B ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download