Intensive care specialists' knowledge,attitudes, and professional use of published research evidence: A mail-out questionnaire survey of appropriate use of research evidence in clinical practice

Purpose

This survey investigates the knowledge, attitudes, and use of published research in clinical practice by intensive care specialists.

Materials and Methods

A mail-out questionnaire was sent to randomly selected intensive care specialists registered with the Australian and New Zealand College of Intensive Care Medicine.

Results

The response rate was 55.9% (133/238). The average score for research knowledge was 2.9 of 6. Eighty-five (65.4%) of 130 respondents reported positive feelings toward using published research evidence in clinical practice, with 96.6% (126/130) reporting use of the concepts of evidence-based medicine at least sometimes. Randomized trials were rated as the most frequently read evidence (rank score, 3.7 of 5), with “Information obtained from the Cochrane Library” the least frequently read (rank score, 2.8 of 5). The most inhibiting barrier to use of published research evidence in practice was “a lack of good evidence providing meaningful answers to clinical problems” (rank score, 3.5 of 5). Eighty-eight (67.7%) of 130 respondents appropriately used published research evidence in clinical practice.

Conclusions

Respondents reported generally positive attitudes toward using published research evidence, in clinical practice; however, room for improvement in technical knowledge relating to published research evidence was noted.     

Assessing the quality of interdisciplinary rounds in the intensive care unit

Purpose

Interdisciplinary rounds (IDRs) in the intensive care unit (ICU) are increasingly recommended to support quality improvement, but uncertainty exists about assessing the quality of IDRs. We developed, tested, and applied an instrument to assess the quality of IDRs in ICUs.

Materials and Methods

Delphi rounds were done to analyze videotaped patient presentations and elaborated together with previous literature search. The IDR Assessment Scale was developed, statistically tested, and applied to 98 videotaped patient presentations during 22 IDRs in 3 ICUs for adults in 2 hospitals in Groningen, The Netherlands.

Results

The IDR Assessment Scale had 19 quality indicators, subdivided in 2 domains: “patient plan of care” and “process.” Indicators were “essential” or “supportive.” The interrater reliability of 9 videotaped patient presentations among at least 3 raters was satisfactory (κ = 0.85). The overall item score correlations between 3 raters were excellent (r = 0.80-0.94). Internal consistency in 98 videotaped patient presentations was acceptable (α = .78). Application to IDRs demonstrated that indicators could be unambiguously rated.

Conclusions

The quality of IDRs in the ICU can be reliably assessed for patient plan of care and process with the IDR Assessment Scale.     

Outcome of hypotensive ward patients who re-deteriorate after initial stabilization by the Medical Emergency Team

Purpose

We evaluated the outcome of hypotensive ward patients who re-deteriorated after initial stabilization by the Medical Emergency Team (MET) in our hospital, due to limited data in this regard.

Methods

One thousand one hundred seventy-nine MET calls in 32184 ward patients from January 2009 to August 2011 were evaluated. Four hundred ten hypotensive patients met study criteria and were divided into: (1)“Immediate Transfers (IT), n = 136”:admitted by MET to intensive care unit (ICU) immediately; (2)“Re-deteriorated Transfers (RDT) n = 72”:initially stabilized and signed off by MET, but later re-deteriorated within 48-hours and admitted to ICU; (3)“Ward Patients (WP) n = 202”: remained stable on ward after treatment.

Results

The RDT and IT had similar APACHE II scores (20.2 ± 5.1 vs. 19.8 ± 4.8; P= .57], but RDT showed hemodynamic stabilization with initial MET resuscitation. Patients who re-deteriorated were younger, took longer for eventual ICU transfer, had higher initial lactic acid and delayed normalization as compared to IT (P < .04). The re-deterioration predominantly occurred within 8-hours of MET evaluation. RDT had higher 28-day mortality than IT and WP; 42% vs. 27% vs. 7% respectively (P < .03). RDT also had a higher rate of endotracheal intubation and worse ICU mortality (P < .01).

Conclusion

Hypotensive ward patients who re-deteriorate after initial stabilization have higher mortality. METs should consider implementing at least an 8-hour follow-up in patients who are deemed stable to remain on the wards after hypotensive episodes.     

Are APLS formulae for estimating weight appropriate for use in children admitted to PICU?

Aim

To determine if the revised APLS UK formulae for estimating weight are appropriate for use in the paediatric intensive care population in the United Kingdom.

Methods

A retrospective observational study involving 10,081 children (5622 male, 4459 female) between the age of term corrected and 15 years, who were admitted to Paediatric Intensive Care Units in the United Kingdom over a five year period between 2006 and 2010. Mean weight was calculated using retrospective data supplied by the ‘Paediatric Intensive Care Audit Network’ and this was compared to the estimated weight generated using age appropriate APLS UK formulae.

Results

The formula ‘Weight = (0.5 × age in months) + 4’ significantly overestimates the mean weight of children under 1 year admitted to PICU by between 10% and 25.4%. While the formula ‘Weight = (2 × age in years) + 8’ provides an accurate estimate for 1-year-olds, it significantly underestimates the mean weight of 2–5 year olds by between 2.8% and 4.9%. The formula ‘Weight = (3 × age in years) + 7’ significantly overestimates the mean weight of 6–11 year olds by between 8.6% and 20.7%. Simple linear regression was used to produce novel formulae for the prediction of the mean weight specifically for the PICU population.

Conclusions

The APLS UK formulae are not appropriate for estimating the weight of children admitted to PICU in the United Kingdom. Relying on mean weight alone will result in significant error as the standard deviation for all age groups are wide.     

No Diversion in Western Massachusetts

Background

Massachusetts (MA) instituted a moratorium on ambulance diversion (“No Diversion”) on January 1, 2009.

Study Objectives

Determine whether No Diversion was associated with changes in Emergency Department (ED) throughput measures.

Design

Comparison of three 3-month periods. Period 1: 1 year prior (January–March 2008); Period 2: 3 months prior (October–December 2008); Period 3: 3 months after (January–March 2009).

Setting

Seven EDs in Western MA; two – including the only Level I Trauma Center – were “high” diversion (≥562 h/year) and 5 were “low” diversion (≤260 h/year). For “all,” “high” diversion and “low” diversion ED groups, we compared mean monthly throughput measures, including: 1) total volume, 2) number of admissions, 3) number of elopements, 4) length of stay for all, admitted and discharged patients. Mean absolute and percent changes were estimated using mixed-effects regression analysis. Linear mixed models were run for “all,” “high” and “low” diversion EDs comparing means of changes between periods. Results are presented as mean change per month in number and percent, and 95% confidence intervals were calculated. We specified that a clinically significant effect of No Diversion had to meet two criteria: 1) there was a consistent difference in the means for both the Period 1 vs. Period 3 comparison and the Period 2 vs. Period 3 comparison, and 2) both comparisons had to achieve statistical significance at p ≤ 0.01.

Results

According to pre-determined criteria, no clinically significant changes were found in any ED group in mean monthly volume, admissions, elopements, or length-of-stay for any patient disposition group.

Conclusion

No Diversion was not associated with significant changes in throughput measures in “all,” “high” diversion and “low” diversion EDs.     

Discordance in interpretation of chest radiographs between pediatric intensivists and a radiologist: Impact on patient management

Purpose

When radiologists are not available, chest radiographs (CXRs) of pediatric intensive care unit (PICU) patients are commonly interpreted by pediatric intensivists. We prospectively investigated the frequency of errors in CXR interpretation by pediatric intensivists and their impact on patient management.

Materials and Methods

Chest radiographs of PICU patients were evaluated by 5 pediatric intensivists then by a pediatric radiologist (the “gold standard”). If the interpretation of the radiologist and intensivist differed, an independent intensivist determined whether a management change took place. A pediatric pulmonologist determined how many intensivist interpretations were different from the radiologist's interpretations.

Results

Seven hundred twenty-eight radiographic findings were identified by the radiologist in 460 CXRs. There were 33 interpretation errors by the intensivists (4.5% of the findings in 7.1% of the CXRs). Only 3/33 error corrections (0.45% of the findings in 0.7% of the CXRs) resulted in change in patient management.

Conclusions

Errors in interpretation of CXRs by pediatric intensivists were common but less than that in other series, probably because of education of the pediatric intensivists through daily rounds with the radiologist. Although interpretation errors that affected patient management were rare, their clinical importance supports the growing practice of 24/7 remote radiograph reading by radiologists.     

Methicillin-resistant Staphylococcus aureus colonization is not associated with higher rate of admission to pediatric intensive care unit

Purposes

Methicillin-resistant Staphylococcus aureus (MRSA) colonization is consistently rising. The question whether the MRSA colonization places the patients at higher risk, requiring higher levels of care when being admitted, has never been studied. We conducted this study to determine the impact of MRSA colonization status on the required level of care upon admission to hospital.

Basic procedures

We conducted a retrospective chart review in 1000 plus–bed tertiary care academic institute. Our study population composed of all the patients who were admitted from January 2011 to March 2011. We found 7413 pediatric admissions that were identified as the study subjects. We assessed and divided study subjects into 2 groups, MRSA colonized and MRSA noncolonized. Methicillin-resistant Staphylococcus aureus–colonized patients were further grouped into those admitted to either pediatric intensive care unit (PICU) or ward, and these 2 groups were analyzed using P value, Fisher exact test, relative risks, and odds ratios.

Main findings

We found a total of 7413 admissions, 753 were admitted in PICU (average pediatric risk of mortality score 18), and 6660 were admitted in pediatric wards (average pediatric risk of mortality score, 5). We found that MRSA colonization was 20 (2.66%) of 753 in PICU admissions and 155 (2.33%) of 6660 in ward admissions. We found that rate of admissions difference between MRSA colonized and MRSA noncolonized groups was clinically insignificant (P > .05).

Principal conclusions

We conclude that MRSA colonization does not increase the need of care in PICU upon admission to hospital from emergency department. However, these preliminary results need to be confirmed through larger, multicenter, and multicountry data analysis.     

Respiratory syncytial virus morbidity, premorbid factors, seasonality, and implications for prophylaxis

Objectives

We investigated factors associated with morbidity and pediatric intensive care unit (PICU) admission in children with respiratory syncytial virus (RSV) infection and explored seasonality and implication of prophylaxis.

Methods

A retrospective study between 2006 and 2008 of every child with a laboratory-confirmed RSV infection was included.

Results

Six hundred seventy RSV admissions were identified. Ten (1.5%) required PICU admissions. Children admitted to PICU were younger than non-PICU admissions (median [interquartile range] age, 0.3 [0.11-0.48] vs 1.18 [0.46-2.49] years; P = .001). Odds associated with PICU admissions included history of chronic lung disease (odds ratio [95% confidence interval], 18.08 [2.29-114.95]; P = .010), history of acyanotic heart disease (7.61 [1.04-42.59], P = .043), and neurodevelopmental conditions (mental retardation, cerebral palsy, or neuromuscular disease; 8.41 [1.63-38.57], P = .012). Odds of bacterial coinfections was 13.50 (1.77-81.29), P = .017. There appeared no significant PICU predilection in terms of sex, history of prematurity, cyanotic heart disease, seizure disorders, chromosomal disorders, or malignancy. Admissions associated with proven RSV infections accounted for 2.4% of PICU annual admissions. The duration of PICU stay was generally brief (median, 3 days). However, median length of hospital stay was significantly longer in the PICU category (8.5 vs 3 days, P < .001). There was no death in the study period. Only 5 (0.75%) of 665 patients were readmitted to the pediatric infectious disease isolation ward in consecutive years, and none required PICU support. Twenty (3%) of admissions involved neonates younger than 30 days. There was no definite seasonality, but incidence was lowest between October and January.

Conclusions

Most infants have mild disease and do not require PICU support. Young infants with history of chronic lung disease, congenital heart disease, and neurodevelopmental conditions appear to be at significantly increased risk for PICU support. There is no winter seasonality for RSV disease in Hong Kong. Therefore, any prophylaxis for at-risk population should provide adequate coverage for the warmer months in subtropical regions.     

An evaluation of instruments for scoring physiological and behavioral cues of pain,non-pain related distress,and adequacy of analgesia and sedation in pediatric mechanically ventilated patients: A systematic review

Objectives

Advancing technology allows for successful treatment of children with life-threatening illnesses. Effectively assessing and optimally treating a child's distress during their stay in the Pediatric Intensive Care Unit (PICU) is paramount. Objective measures of distress in mechanically ventilated pediatric patients are increasingly available but few have been evaluated. The objectives of this systematic review were to identify available instruments appropriate for measuring physiological and behavioral cues of pain, non-pain related distress, and adequacy of analgesia and sedation in mechanically ventilated pediatric patients, and evaluate these instruments in terms of their psychometric properties.

Design

A systematic review of original and validation reports of objective instruments to measure pain and non-pain related distress, and adequacy of analgesia and sedation in mechanically ventilated PICU patients was undertaken.

Data sources

A comprehensive search was conducted in 10 databases from January 1970 to June 2011. Reference lists of relevant articles were reviewed to identify additional articles.

Review methods

Studies were included in the review if they met pre-established eligibility criteria. Two independent reviewers reviewed studies for inclusion, assessed quality, and extracted data.

Results

Twenty-five articles were included, identifying 15 instruments. The instruments had different foci including: assessing pain, non-pain related distress, and sedation (n = 2); assessing pain exclusively (n = 4); assessing sedation exclusively (n = 7), assessing sedation in mechanically ventilated muscle relaxed PICU patients (n = 1); and assessing delirium in mechanically ventilated PICU patients (n = 1). The Comfort Scale demonstrated the greatest clinical utility in the assessment of pain, non-pain related distress, and sedation in mechanically ventilated pediatric patients. Modified FLACC and the MAPS are more appropriate, however, for the assessment of procedural pain and other brief painful events. More work is required on instruments for the assessment of distress in mechanically ventilated muscle relaxed PICU patients, and the assessment of delirium in PICU patients.

Conclusions

This review provides essential information to guide PICU clinicians in choosing instruments to assess pain, non-pain related distress, and adequacy of analgesia and sedation in mechanically ventilated pediatric patients. Effective knowledge translation is essential in the implementation, adoption, and successful use of these instruments.     

Hospitalist bed management effecting throughput from the emergency department to the intensive care unit

Rationale

Emergency department (ED) patients in need of an intensive care unit (ICU) admission are very sick. Reducing the length of time to get these patients into ICU beds is associated with improved outcomes.

Objective

To reduce the ED length of stay for patients requiring admission to the medical ICU or coronary care unit through the implementation of the “active bed management” (ABM) intervention.

Methods

A pre-post study design compared data from November 2006 to February 2007 with those from those same months in the prior year at Johns Hopkins Bayview Medical Center in Baltimore. The ABM intervention was carried out by hospitalist physicians and involved: (i) making triage decisions for patients to be admitted and facilitating their transfer from ED to the appropriate care setting and (ii) having proactive management of Department of Medicine resources, which included twice-daily ICU bed management rounds and regular visits to the ED to assess flow.

Measurement

Throughput time for patients presenting to the ED requiring ICU admission was analyzed.

Main Results

The ED census was higher during the intervention period as compared with the control period, 17?573 versus 16?148 patients. Throughput from ED to coronary care unit and medical ICU beds was reduced by 99 (±14) minutes (from 353 minutes in the control period to 254 minutes in the 4 months after the initiation of ABM, P < .0001). Staffing, length of stay, case mix index, ICU transfer rates, and ICU death rates were stable across the 2 periods, all P = not significant.

Conclusion

Conscientious management of hospital beds, in this case by hospitalist physicians providing ABM, can have a positive and substantial impact on the ED throughput of critically ill patients admitted to ICU beds. This efficiency is likely to positively have impacted on patient satisfaction and safety.     

Assessing Locomotion Ability in West African Stroke Patients: Validation of ABILOCO-Benin Scale

Objective

To calibrate and validate the Benin version of ABILOCO, a Rasch-built scale developed to assess locomotion ability in stroke patients.

Design

Prospective study and questionnaire development.

Setting

Rehabilitation centers.

Participants

Stroke patients (N=230; mean age ± SD, 51.1±11.6y; 64.3% men).

Intervention

Not applicable.

Main Outcome Measures

Participants completed a preliminary list of 36 items including the 13 items of ABILOCO. Items were scored as “impossible,” “difficult,” or “easy.” The mobility subdomain of FIM (FIM-mobility), the Functional Ambulation Classification (FAC), the 6-minute walk test (6MWT), and the 10-meter walk test (10MWT) were used to evaluate and elucidate the validity of the ABILOCO-Benin scale.

Results

Successive Rasch analyses led to the selection of 15 items that define a unidimensional, invariant, and linear measure of locomotion ability in stroke patients. This modified version of the ABILOCO scale, named ABILOCO-Benin, showed an excellent internal consistency, with a Person Separation Index of .93, and excellent test-retest reliability with high intraclass correlation coefficients of .95 (P<.001) for item difficulty and .93 (P<.001) for subject measures. It also presented good construct validity compared with FAC, FIM-mobility, 6MWT, and 10MWT (r≥.75, P<.001).

Conclusions

ABILOCO-Benin presents good psychometric properties. It allows valid, reliable, and objective measurements of locomotion ability in stroke patients.     

Developing and evaluating a machine learning based algorithm to predict the need of pediatric intensive care unit transfer for newly hospitalized children

Background

Early warning scores (EWS) are designed to identify early clinical deterioration by combining physiologic and/or laboratory measures to generate a quantified score. Current EWS leverage only a small fraction of Electronic Health Record (EHR) content. The planned widespread implementation of EHRs brings the promise of abundant data resources for prediction purposes. The three specific aims of our research are: (1) to develop an EHR-based automated algorithm to predict the need for Pediatric Intensive Care Unit (PICU) transfer in the first 24 h of admission; (2) to evaluate the performance of the new algorithm on a held-out test data set; and (3) to compare the effectiveness of the new algorithm's with those of two published Pediatric Early Warning Scores (PEWS).

Methods

The cases were comprised of 526 encounters with 24-h Pediatric Intensive Care Unit (PICU) transfer. In addition to the cases, we randomly selected 6772 control encounters from 62516 inpatient admissions that were never transferred to the PICU. We used 29 variables in a logistic regression and compared our algorithm against two published PEWS on a held-out test data set.

Results

The logistic regression algorithm achieved 0.849 (95% CI 0.753–0.945) sensitivity, 0.859 (95% CI 0.850–0.868) specificity and 0.912 (95% CI 0.905–0.919) area under the curve (AUC) in the test set. Our algorithm's AUC was significantly higher, by 11.8 and 22.6% in the test set, than two published PEWS.

Conclusion

The novel algorithm achieved higher sensitivity, specificity, and AUC than the two PEWS reported in the literature.     

Counselors' reflections on the administration of screening and brief intervention for alcohol problems in the emergency department and 3-month follow-up outcome

Objectives

The purpose of this paper was to explore how events that counselors endorsed occurring during an emergency department–based screening and brief intervention (SBI) for drinking discriminate patients who reported change in Alcohol Use Disorder Identification Test (AUDIT) domains at follow-up from those who did not.

Method

Patients who scored “>5” on the AUDIT were eligible for SBI. At the end of each intervention, counselors completed the questionnaire indicating which parts of the intervention they just used.

Results

Discriminant function analyses indicated that “Referral made” discriminated for alcohol intake change (Wilks' λ = 0.993, P < .05); “Did the patient set goals during intervention?” and “Referral made” discriminated for alcohol dependency change (Wilks' λ = 0.940 and Wilks' λ = 0.919, P < .05, respectively). “Intention to quit” (Wilks' λ = 0.984, P < .05) discriminated for alcohol-related harm change.

Conclusions

Making referrals to addiction treatment during motivational intervention discriminated for alcohol intake and dependency change. Working on intention to quit is an important point in changing alcohol-related harm. When conducting the SBI in ED, counselors may be mindful in making appropriate referrals to address alcohol use and examine intention to quit to maximize the efficacy of the harm-reduction approach.     

Effect of interhospital transfer on resource utilization and outcomes at a tertiary pediatric intensive care unit

Purpose

The study aimed to examine the effect of interhospital transfer on resource utilization and clinical outcomes at a tertiary pediatric intensive care unit (PICU) among patients with sepsis or respiratory failure.

Materials and methods

Data on 2146 consecutive admissions with respiratory failure or sepsis to the PICU were analyzed. Data included demographics, admission source, and outcomes. Admission source was classified as interhospital transfer from the emergency departments (ED), wards, or PICUs of referring hospitals; or from the study hospital ED (direct).

Results

Compared with direct admissions, inter-PICU transfers had higher crude mortality (odds ratio, 1.93; 95% confidence interval, 1.31-2.84) but not significant mortality difference (odds ratio, 1.16; 95% confidence interval, 0.71-1.86) after adjusting for illness severity, age, and sex. Conversely, ED transfers had lower PICU mortality than direct ED admissions. Children with transfer admissions stayed significantly longer and used more intensive care technology in the study PICU than children directly admitted (P < .01). In comparisons within quartiles of mortality risk, inter-PICU transfers had longer hospitalization and higher mortality in all but the highest quartile.

Conclusions

Interhospital transfer, particularly inter-PICU transfer, was associated with significant hospital resource consumption that often correlated with admission illness severity. Future prospective studies should identify determinants of pretransfer illness severity and investigate decision making underlying interhospital transfer.     

Functional Outcomes Associated With Adaptive Seating Interventions in Children and Youth With Wheeled Mobility Needs

Objective

To determine the parent-reported functional outcomes associated with adaptive seating devices for wheeled mobility devices used by young people aged 1 to 17 years.

Design

Longitudinal case series.

Setting

Homes of participating parents.

Participants

Parents (N=70, 63 mothers, 6 fathers, 1 grandmother) who had children with adaptive seating needs.

Intervention

Adaptive seating system for wheeled mobility devices.

Main Outcome Measure

Family Impact of Assistive Technology Scale for Adaptive Seating (FIATS-AS).

Results

All parents completed the FIATS-AS 4 times—2 times before and 2 times after their child received a new adaptive seating system. Mixed-design analysis of variance did not detect significant mean differences among the FIATS-AS scores measured at baseline and 2 and 8 months after receiving the seating system (F_2,134=.22, P=.81). However, the FIATS-AS detected a significant interaction between age cohort and interview time (F_4,134=4.5, P<.001, partial η²=.16). Post hoc testing confirmed that 8 months after receiving the seating system was associated with a large improvement in child and family functioning for children <4 years, maintenance of functioning for children between 4 and 12 years, and a moderate decline in functioning for youth between 13 and 17 years.

Conclusions

Adaptive seating interventions for wheeled mobility devices are associated with functional changes in the lives of children and their families that interact inversely with age. Future controlled longitudinal studies could provide further empirical evidence of functional changes in the lives of children and their families after the introduction and long-term use of specific adaptive seating interventions.     

Continuous end-tidal carbon dioxide monitoring in pediatric intensive care units

Purpose

End-tidal carbon dioxide (ETCO₂) monitoring has a variety of clinical applications in critically ill pediatric patients. This study was designed to explore the current availability and utilization patterns for continuous ETCO₂ monitoring in pediatric intensive care units.

Methods

A Web-based survey was distributed to directors of all accredited pediatric critical care fellowship programs in the United States.

Results

Sixty-six percent of directors completed this survey. One hundred percent of directors had access to ETCO₂ monitoring for intubated patients and 57% for nonintubated patients. Eighty-three percent of respondents used ETCO₂ monitoring “always” or “often” for endotracheal tube confirmation. Fifty percent of respondents used ETCO₂ monitoring “always” or “often” for cardiopulmonary resuscitation, 38% for moderate sedation, and 5% for acid-base disturbances. All respondents who used ETCO₂ monitoring felt that it was easy to use. The most common reason for not using ETCO₂ monitoring was lack of availability (75%).

Conclusions

End-tidal carbon dioxide monitoring is widely available and used for intubated patients. However, it could be applied more frequently in other clinical situations in pediatric intensive care units.     

A history of smoking is associated with improved survival in patients treated with mild therapeutic hypothermia following cardiac arrest

Objectives

To assess the association between smoking and survival with a good neurologic outcome in patients following cardiac arrest treated with mild therapeutic hypothermia (TH).

Methods

We conducted a retrospective observational study of a prospectively collected cohort of 188 consecutive patients following cardiac arrest treated with TH between May 2007 and January 2012. Smoking status was retrospectively collected via chart review and was classified as “ever” or “never”. Primary endpoint was survival to hospital discharge with a good neurologic outcome and was compared between smokers and nonsmokers. Logistic regression analysis was used to assess the association between smoking status and neurologic outcome at hospital discharge; adjusting for age, initial rhythm, time to return of spontaneous circulation (ROSC), bystander CPR, and time to initiation of TH.

Results

Smokers were significantly more likely to survive to hospital discharge with good neurologic outcome compared to nonsmokers (50% vs. 28%, p = 0.003). After adjusting for age, initial rhythm, time to ROSC, bystander CPR, and time to initiation of TH, a history of smoking was associated with increased odds of survival to hospital discharge with good neurologic outcome (OR 3.54, 95% CI 1.41–8.84, p = 0.007).

Conclusions

Smoking is associated with improved survival with good neurologic outcome in patients following cardiac arrest. We hypothesize that our findings reflect global ischemic conditioning caused by smoking.     

Unplanned pediatric intensive care unit readmissions: A single-center experience

Purpose

The purpose of the study was to compare patients readmitted to the pediatric intensive care unit (PICU) unexpectedly within 48 hours (early), more than 48 hours from transfer (late), or not readmitted during the same hospitalization.

Materials and Methods

A retrospective study (2007-2009) was performed at a tertiary care pediatric academic hospital. Readmitted at-risk patients were grouped by timing of readmission, and a sample of nonreadmitted patients was randomly selected. Early readmissions were compared to late readmissions and to nonreadmissions.

Results

Of 3805 eligible patients, 3.9% had an unplanned PICU readmission with almost half occurring within 48 hours. Median times to readmission were 21.5 hours (early) and 7 days (late). Compared with late readmissions, early readmissions were more often admitted from and transferred to a surgical service, transferred on a weekend, and readmitted with the same primary diagnosis. Compared with nonreadmitted patients, independent risk factors for early readmission were admission source and respiratory support at PICU transfer. Readmitted patients had longer total PICU and hospital lengths of stay than nonreadmitted patients. Late readmissions had a higher mortality than early readmissions.

Conclusions

Patients requiring an unplanned PICU readmission had worse outcomes than those without a readmission. Future studies should focus on identifying modifiable risk factors for targeted interventions.