Specialist Dissertation Services
Dissertation Examples

Chapter 2

Literature Review

Chapter 2 presents the literature review to evaluate the current state of the science as it relates to the hemodynamic effects assessed through EC of mechanical ventilation on premature infants targeting gestational ages less than 37 weeks. The focus of this Chapter is to examine the existing body of science and situate the proposed research question within the current body of literature. Research findings are analyzed and synthesized across multiple related research studies. Hemodynamic measurements may demonstrate important statistically significant correlation or predictive value in this fragile population.

Review of the Literature

Chapter 2 incorporates the search strategy, key findings, limitations, and themes from the literature, and synthesizes the data to provide the structure for the chapter.  Neonatal research in noninvasive technology to monitor the status of premature infants is continually expanding and evolving.  The search strategy for this study is ongoing until the research is completed. Chapter 2 contains five sections that investigate the independent variable mechanical ventilation, the dependent variables derived through hemodynamic monitoring with gestational age, and covariates that may influence hemodynamic measurements.  The first section outlines the literature review strategy, conceptualizes current practice through the lens of changes in the last decade affecting cardiopulmonary status, and categorizes current, relevant empirical literature within the scope of the study.  The second section reviews current advances affecting hemodynamic status including gestational age, anemia, and erythrocytes endowment.  The third section presents the research related to the premature lung including surfactant theory, gentler modes of ventilation, mechanical ventilation and the effects of oxygen therapy, stretch, and inflammation.  The fourth section looks at the hemodynamic effects of mechanical ventilation and current nursing assessment of hemodynamic status.  The fifth section reviews the current state of the literature on EC including normative data, comparison to echocardiography, and advantages from the literature including reported limitations.  There were no prior studies found that examined the specific questions in this investigation so the literature review is situated within current practice.  Chapter 2 ends with a summary of the literature review.

Literature Review Strategy

A comprehensive and systematic literature search was conducted to determine the current empirical reality and state of knowledge about the hemodynamic effects assessed through EC of mechanical ventilation on premature infants including available data related to gestational age.  Elimination of redundant records, document storage, and data manipulation used ProCite and RefWorks depending on the institution used for the search.  The search yielded 3865 citations (Table 1).  The initial selection of articles for this review took place October 22, 2016, and November 10, 11, and 13, 2016, guided by the question: “What are the hemodynamic effects assessed through EC of mechanical ventilation on premature infants in a Level III neonatal intensive care unit controlling for gestational age.”

The computer-based literature search included studies between the years 2011 through 2017, with a few important earlier studies to assess the state of the literature.  The search strategy includes five electronic databases including Cumulative Index to Nursing and Allied Health (CINAHL), EBSCO Federated Search, Web of Science, Scopus (which covers all journals in PubMed/Medline and EMBASE from 1996), ProQuest, and hand searched abstracts from the Society of Pediatric Research.  To establish the search Medical Subject Headings (MeSH) terms, and keywords were created.  A combination of controlled vocabulary, words, and phrases were included in the search strategy.  MeSH terms included electric impedance; hemodynamics; ventilators, mechanical; limits: age groups, inclusion of infant, premature; publication types now including dissertations and theses; and clinical trial.  Filters placed included full text, English language, and peer reviewed.  ProQuest searched dissertations and theses within the population.  The search strategy included wildcard characters (*), proximity operators (“”) and search operators (AND, OR).  The title/abstract search included premature infant AND intensive care unit neonatal AND/OR mechanical ventilation AND/OR electrical cardiometry, electrical velocimetry, thoracic electrical bioimpedance AND/OR hemodynamic.

Search Strategy

The search strategy to explore and refine the search included:

1. Neonate OR neonatal OR newborn OR newborns OR infants OR premature OR premature infant OR paediatric
2. Cardiac output OR heart output OR cardiac index OR heart Index OR stroke volume OR stroke volume index
3. Hemodynamic* AND measure* OR monitor* OR determination
4. #1 OR #2
5. #1 OR #3
6. # 1 OR #2 OR #3
7. Mechanical ventilation OR respirator OR ventilation OR vent*
8. Electrical cardiometry OR bioimpedance OR thoracic electrical bioimpedance OR electrical velocimetry
9. Packed red blood cells
10. Limits of viability
11. # 5 OR #6
12. # 1 OR #7
13. #2 OR #7
14. #1 OR #8
15. Letter OR editorial OR news OR comment OR case reports OR conference OR dissertation OR review OR meta synthesis OR note OR poster OR theses
16. Intensive care unit neonatal AND NICU AND newborn ICU
17. Delayed cord clamping OR DCC OR umbilical cord clamping OR UCC
18. #1 OR #15
19. #3 OR # 15

Study selection included screening the titles of the articles and the abstracts for relevance to the research question.  Relevant abstracts were selected for a full review and those remaining were selected for the synthesis.  Articles were eligible if papers were quantitative research published, peer-reviewed, and indexed in scientific journals, conferences, podium presentations, posts, or books.  Exploration, analysis, and a summary of the pertinent literature are included in Table 1 and the final analysis in Table 2.

Table 1:  Databases statistics for the literature review.
Database	Found	Excluded	Analyzed	Synthesized
CINAHL	2120	2100	8	2
EBSCO	1025	1021	8	3
Scopus	416	400	9	3
Web of Science	101	136	8	1
Total	3662	367	33	9

Table 2:

Flow Diagram of the Literature Search Process

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Factors Affecting Hemodynamic Status

Limits of viability

The current limits of viability for premature is 23 weeks and 500 grams with some centers resuscitating infants as young as twenty weeks gestation (American Academy of Pediatrics, 2015).  Increasing gestational age provides maturational advantages that reduce morbidities and mortality (American Academy of Pediatrics, 2015).  Serenius, Kallen, and the Express Group (2013) identified significant poor developmental outcomes in premature infants that were inverse to gestational age including seventy-eight percent with severe developmental delays, twelve percent with cerebral palsy, and almost ten percent with blindness.  A recent publication by Joseph et al. (2016) examined neurocognitive outcomes in 889 children at ten years of age born at less than 28 weeks gestation investigating neurocognitive outcomes.  Over half of these children suffered moderate to severe deficits and have persistent impairments in academic, social, physical, cognitive, and behavioral performance.  Rhee et al. (2016) looked at 185 infants less than 33 weeks gestation being mechanically ventilated investigating the incidence of severe intraventricular hemorrhage finding alterations in diastolic cerebral blood flow velocity and dysregulation inverse to gestational age.  Smaller premature infants showed greater risk of cerebral dysregulation and increased risk of hemorrhage.

These recent studies were robust and well powered looking at the impact of short- and long-term outcomes of mechanically ventilated premature infants.  These findings underscore the critical need for approaches to care that monitor and support the new frontier of premature infants.  Survival from the neonatal intensive care unit is no longer the acceptable goal.

Anemia and Hemodynamics

Anemia in the neonatal intensive care unit is a common problem and influenced by gestational age, day of life, underlying pathology, fetal hemoglobin, and blood sampling (Ibrahim, Ying Ho, & Yeo, 2014).  Premature infants require more red blood cell transfusions that any other hospitalized population (Shereen, Yasmeen, Ibrahim, & Mohamed, 2012).

In reviewing the current literature on cardiovascular performance and packed red blood cell transfusion, several studies have investigated the effect of packed red blood cell transfusion on cerebral blood velocities, perfusion index, oxygenation, and cardiac output.  These studies have investigated cardiovascular function pre and post transfusion using ultrasound and echocardiography.  In one prospective observational study of 35 premature infants, cardiac output was assessed using an ultrasound device to study the effect of transfusion on perfusion index and cardiac out (Kanmaz, et al., 2013).  Cardiac output and heart rate increase in response to anemia.  Another study of 103 premature infants using echocardiography assessed myocardial performance in anemic premature infants and found an inverse relationship between pre and post transfusion left ventricular function (Radicioni, Troliani, & Mezzetti, 2012).  Saleemi and colleagues investigated cardiac functioning using echocardiography in very low birth weight infants pre and post transfusion and found cardiac functioning improves following transfusion (Saleemi, et al., 2013). Quante and colleagues investigated heart rate, cardiac output, and cerebral blood flow after transfusion finding decreases in these parameters (Quante, et al., 2013).  Hemodynamically anemia has an effect on cardiopulmonary status and hemoglobin level is included as a covariate in this study.

The criteria for transfusion in the premature infant are subjective, non-standardized, vary among institutions and practitioners, and are not driven by strong clinical indicators or empirical data (Quante, et al., 2013; Ibrahim, Ying Ho, & Yeo, 2014).  There is limited evidence on necessary transfusion thresholds, clinical indicators of which infants will benefit most from transfusion, long-term neurodevelopmental outcomes of transfused infants, and transfusion-related incidents.  These studies are robust with small to moderate effect sizes.  Across all studies, anemia affects hemodynamic status and transfusion produces notable changes in systemic blood flow, heart rate, oxygenation, and right ventricular filling pressures.  Limitations of the studies include different cut points for transfusion, the transfusion of varying amounts of packed blood cells, and unstandardized lengths of transfusion.  These findings led into a recent change in practice addressing transfusion and anemia bringing into question the effect of delayed cord clamping (DCC), immediate cord clamping (ICC), and umbilical cord milking (UCM) on hemodynamic measurements.

Umbilical Cord and Hemodynamics

A recent change in practice that benefits the pulmonary and cardiovascular system as well as hemodynamics is DCC and UCM.  DCC and UCM provide an additional endowment of erythrocytes at birth to the premature infant reducing the risk of anemia, intraventricular hemorrhage, transfusion, and oxygen therapy at 36 weeks (Alan, et al., 2014; Erikson-Owens, Mercer, & Oh, 2012; Katheria & Leone, 2013; Takami, et al., 2012).  DCC and UCM increase both iron stores and increase blood volume (Katheria & Leone, 2013).  Seven studies in the last five years looked at DCC or UCM versus immediate cord clamping (ICC), and of those, only two studies (n = 260) investigated the hemodynamic effects of this procedure on hemodynamic measurements.  Katheria et al. (2013) randomized 60 preterm infants less than 31 weeks gestation to evaluate UCM (n = 30) versus immediate cord clamping (n = 30) and systemic blood flow in the first 6 to 30 hours of life using three blinded serial echocardiograms.  Hemodynamic benefits in the UCM group included higher right ventricular outflow and superior vena cava flow through 30 hours of life.  Upadhyay et al. (2013) randomized 200 infants less than 35 weeks gestation to UCM (n = 100) and ICC (n = 100) and similarly used echocardiography to study the effects of UCM at three-time points.  The authors found mean blood pressure and heart rate to be within normal range for the infant and improved diastolic function increasing left ventricular preload.  Although not highlighted, both teams measured heart rate and blood pressure and indicated there were no significant differences.  However, both found large statistical differences using echocardiography in preload, afterload, and contractility measurements.  Transfusion appears to have an effect on both the preterm and near-term infants hemodynamic function not noticeable by routine vital sign monitoring but noticeable by echocardiography.

In summary, the recent changes in practice investigated the use of UCC and DCC versus ICC, and two studies looked at hemodynamic changes.  The existing literature demonstrates benefit in elevated blood counts equating to lower transfusion needs, and reduced need for resuscitation.  These studies do not investigate the impact on morbidity and mortality.  The populations studied were dissimilar, reported conflicting results, and different gestational age groups.  The neonatal intensive care unit serves populations, within populations and new technologies must be capable of monitoring different levels of maturation or investigated as to where they are most useful.  Gaps in the literature are a lack of large randomized samples to investigate; the optimal time to clamp the cord, the best length needed for transfer of red blood cells, the impact of various cord sizes, the effects on infants at various gestational ages, and which infants benefit most from the intervention.  There is limited data on the effects of random clamping on the cardiopulmonary system and hemodynamic measurements.  Further research is necessary to optimize this therapy in the delivery room.

Table 3:  Umbilical Cord Literature and Hemodynamic Effect (n = 571)
Author	Year	Sample size	Gestational age	Weight	Control	Number of times	Speed  (seconds)	Hemodynamic Effect
Alan, et al.	2014	44	< 32 weeks	< 1500 grams	ICC	3	5 cm	Used heart rate and noninvasive blood pressure
Erikson-Owens, et al.	2012	24	Term	< 1200 grams	ICC	5		Used heart rate
Katheria, et al.	2014	60	23-31 weeks and 6 days	1200-1810 grams	ICC	3	20cm within 2s	Improved systemic blood flow, higher preload
March et al.	2013	75	24-28 weeks	< 1500 grams	ICC	3	NA
Rabe, et al.	2012	58	24-32 weeks 6 days	1200-1950 grams	DCC	4	20 cm/ 2 s
Takami, et al.	2012	50	< 29 weeks	< 1250 grams	ICC	3	10 cm
Upadhyay, et al.	2013	200	> 35 weeks	< 1500 grams	None	3	10 cm	Improved cardiac output

 

Protecting the Premature Lung

Recent advances in the care of the premature infants have significantly improved survival and changed the approach to respiratory care to a proactive, minimally invasive approach, aimed at improving survival without profound impairment (Flannery, et al., 2016).  This shift in approach has resulted in fewer infants requiring mechanical ventilation in some gestational ages but accompanying this shift is a population of premature infants at the lower limits of viability.  These infants at the lowest extremes suffer the complications of mechanical ventilation at staggering and significant proportions.  There are numerous reasons for the application of mechanical ventilation in premature infants; the most prevalent is respiratory distress (RDS) and failure (Flannery, et al., 2016).

Early surfactant therapy

One of the most significant advances in the treatment of respiratory distress in the premature infant is the administration of antenatal steroids to pregnant women threatening to deliver early and the administration of exogenous surfactant (American Academy of Pediatrics, 2015).  Exogenous surfactant is a synthetic protein administered intra-tracheal to reduce surface tension, improve pulmonary compliance, and prevent lung collapse (Katheria & Leone, 2013).  Surfactant has greatly improved survival rates for infants born less than 28 weeks gestation in the United States (Katheria & Leone, 2013).  However, with this survival has come increased morbidity including chronic lung disease termed bronchopulmonary dysplasia (BPD), poor neurodevelopmental outcomes, blindness, and intraventricular hemorrhage (American Academy of Pediatrics, 2015; Flannery, et al., 2016).

The literature review identified four studies that looked at surfactant administration and hemodynamics using echocardiography with a focus on the effect on the PDA (Table 4).  The PDA is an intrauterine shunt or bypass that redistributes pulmonary blood flow until birth but can sustain delays in closure in the premature infant.  Vitali et al.(2014) conducted a prospective observational study looking at hemodynamic changes in 14 infants less than 34 weeks gestation treated with surfactant within the first 72 hours of life using Tricuspid Annular Plane Systolic Excursion (TAPSE).  TAPSE is an echocardiographic measurement that assesses right heart function not typically used in the neonatal population.  The researchers noted improvements in right heart function after surfactant administration attributed to preload reduction.  Fuji, Allen, Doros, and O’Brien (2010) had similar findings but used standard transthoracic echocardiography techniques (TTE).  Katheria and Leone (2012) evaluated hemodynamic status after surfactant administration in twenty infants using TTE and found no significant changes in the PDA, and increased systemic blood flow.  Sehgal et al. (2010) looked at delivery room surfactant treatment and reported a drop in arterial pressure, increase in left-to-right flow through the PDA, a decrease in left ventricular output, and an increase in right ventricular output.

Table 4:  Surfactant Administration and Echocardiographic Hemodynamic and PDA Changes
Author	Date	Gestational age	Age	Weight	Sample	PDA by Echo	Hemodynamics
Fuji, et al.	2010	27.1 + 1.6 weeks	< 24 hours of age	930 + 231 grams	50	Reduced left to right shunt through PDA	Large reduction in the PDA, lower right ventricular pressure, and lower systolic arterial pressure
Katheria, et al.	2013	28.3 ± 2	< 24 hours of age	1160 ± 133 grams	20	No change PDA	Increased systemic blood flow and no change PDA
Sehgal et al.	2010	28.3 + 1.3	Within 30 minutes of birth	1289 + 224 grams	16	Increased shunt through PDA	Increased right heart pressures but decreased left heart pressure
Vitali, et al.	2014	27.5-33.4	< 72 hours age	1162 grams (500-1990)	14	Reduction in shunt of PDA	No change contractility, increased right heart pressures

The demographics of today’s neonatal intensive care unit are changing and the infants are much more premature, critically ill, and in need of highly technical and skilled care.  The high-risk premature infant at the limits of viability has unique oxygenation and perfusion challenges never faced before and the published literature is conflicting.  There were no published studies using EC.  The evidence on the hemodynamic effects of surfactant by echocardiography related to the PDA varied from no change in hemodynamics to substantial changes in left and right heart function.  Premature infants were all below 34 weeks gestation and similar in weight and day of life but the sample sizes were small.  The inconsistent findings regarding the hemodynamic effects of surfactant require further research.

Gentler modes of ventilation

The majority of infants born prior to 30 weeks gestation will require some level of respiratory support (Chang, et al., 2016).  Advances in the respiratory management of the premature infant have led to what is termed gentler modes of ventilation to avoid the exposure to the risks and adverse effects associated with the more invasive mechanical ventilation.  Noninvasive means of ventilation include nasal intermittent positive pressure ventilation (NIPPV), high flow nasal cannula (HFNC), and nasal continuous positive airway pressure (NCPAP).  These respiratory treatment modalities’ primary purposes are to address respiratory distress and act as a bridge during weaning from mechanical ventilation.  These newer treatment modalities are not risk-free and have the potential to cause air leaks, bowel perforation, gastric distention, nasal septal trauma, pneumothorax, and respiratory failure requiring intubation (Wilkinson, et al., 2016).

Chang and colleagues (2016) investigated the effects of NCPAP and NIPPV on 20 stable preterm infants with a mean gestational age of 27 weeks and the hemodynamic effects assessed with ultrasonography comparing each infant to themselves for 30 minutes followed by the alternative treatment.  The authors found no statistical differences between the groups when evaluating ventricular output, superior vena cava flow, and cerebral blood flow velocity.  Wilkinson et al. (2016) conducted a Cochrane Systematic Review of 15 trials to compare the safety and effectiveness of noninvasive respiratory support as a viable alternative for stable premature infants.  The authors concluded that all forms of noninvasive treatment had similar efficacy in preventing reintubation, death, and chronic lung disease, but premature infants still require mechanical ventilation especially the very low birth weight.  Although an older study, the most recent work looking at infants less than 29 weeks in the literature was by Mortiz, Fritz, Mann, and Simma (2008) who investigated the hemodynamic effects of NCPAP on cardiac output using echocardiography and found no difference in cardiac output in twenty-one stable preterm infants with a mean gestational age of 28 weeks.

The published research on the gentler forms of ventilation are limited, but congruent and similar to earlier findings reviewed in 2000 to 2010 however, the sample sizes are small and narrow in the gestational age ranges.  Gentler forms of ventilation in specific populations do not appear to have a significant effect on hemodynamics in stable premature infants less than 28 weeks gestation, but infants requiring mechanical ventilation across wider gestational age groups and those who are unstable require further analysis.

Mechanical Ventilation

Approximately half of infants born at 23 to 28 weeks require intubation and mechanical ventilation to support oxygenation and ventilation (American Academy of Pediatrics, 2015).  Currently, a multitude of commercial ventilators are available, and many neonatal intensive care units require multiple types to meet the varied needs of every type of respiratory disease the current neonatal provider faces (Polin, 2012).  The newer equipment is sophisticated and technical requiring education and experience to support clinical expertise and application.  With the newer sophistication and technology come complexity, risk, safety concerns, and lack of research and theory to guide practice.  The published research on mechanical ventilation documents risk that directly affects hemodynamic status including the effects of too much or too little oxygen, pressure, and volume trauma, as well as inflammation.

Hyperoxia 

Too much oxygen or hyperoxia in the premature infant is harmful and well documented.  Hyperoxia, defined as a saturated pulse oximetry above 92% is reported to contribute to changes in the alveoli, smooth muscle of the airway, remolding of the premature lung, cytokine, and protein expression (Vakrolova, et al., 2015; Tan, et al., 2016).  Recent research by Tan and colleagues (2016) identified reduced expression of silent information regulator 1 (SIRT 1) during exposure to three levels of oxygen and the incidence of chronic lung disease in 28 to 30 weeks gestation premature infants.  The authors found that increased exposure to oxygen decreased expression of this protein contributing to the BPD, which is chronic lung disease of the premature.  Vakrolova et al. (2015) found significant changes to lung tissue, brain injury, and vascular damage to the retina from hyperoxia in 683 very low birthweight infants.  The authors found that the frequency of BPD correlates with gestational age and birthweight.  These researchers concluded that oxygen-induced injuries not only affect the length of stay and co-morbid conditions in the neonatal intensive care unit, but also contribute to vascular disease and influence cardiopulmonary function longitudinally.

These studies add support to the short- and long-term consequences of oxygen usage and the consequent development of BPD in the premature infant.  Both studies contain large samples with large effect size and adequate power although were not randomized.  Current molecular biology and genetics research are beginning to explain how the alterations in oxygen exposure have detrimental effects on transcriptional mechanisms necessary for normal lung architecture and growth.  Mechanical ventilation alters the developing lung via a variety of measures including mechanical stretch, oxygen exposure, and inflammation.   Factors that affect the lungs contribute to stress on the cardiovascular system.

Hypoxia

In premature infants, hypoxemia is a saturated pulse oximetry reading below 85% (Vijlbrief, et al., 2014).  The literature contains multiple studies investigating the morphological changes of the lung in the premature infant and a multitude of changes that include thickening of the smooth muscles of the airway, inflammation, changes in mucous production, and alterations in growth trajectory (Keglowich, Baraket, Tamm, & Borger, 2014).  Hartman et al. (2012) and Keglowich et al. (2014) investigated the effects of oxygen administration on lung morphology, signaling mechanisms, and molecular physiology using fetal airway smooth muscle cells in 18 to 20 week gestation premature infants.  These scientists identified reduced oxygen affects the rapidly developing smooth muscle of the airway and the expression of proinflammatory markers.

Fluctuations and extremes in oxygen levels have profound effects on the premature lung and organogenesis.  Studies looking at both extremes of oxygen administration are robust, well documented, and of pertinent to neonatal nurses.  The multi-faceted nature of the care and management of the premature infant has contributed to a better understanding but has not reduced the incidence of poor outcomes.  These authors point out that the detrimental effects of oxygen exposure on organogenesis have not only short, but also long-term implications on organ development, remodeling, and quality of life.

Mechanical Stretch and Inflammation

Mechanical ventilation in the premature infant creates changes in the lungs via intrapleural and intrathoracic pressure or volume alterations that can affect the cardiovascular system (Fajardo, et al., 2014; Seghal, et al., 2010).  Changes within the thorax affect filling of the right and left ventricle with blood, the heart’s contractility, and the work the heart must overcome to eject the blood.  Mourani and colleagues (2014) investigated the bronchopulmonary dysplasia (chronic lung disease in premature infants) and pulmonary hypertension (increased pressure in the arteries of the lungs) using echocardiography in 247 infants at seven days of life and 36 weeks gestation.  Risk factors for lung disease included lower gestational age, lower birth weight, multiple gestation pregnancies, and positive pressure ventilation.  Bohrer, Silveira, Neto, & Procianoy (2010) studied 19 premature infants with a mean gestational age of 35.8 weeks and plasma levels of inflammatory markers after 2 hours of mechanical ventilation.  They found that a minimal period of exposure to mechanical ventilation, approximately 12 positive pressure breaths, initiates inflammation, which signals cellular inflammatory markers that disrupt normal growth and contribute to remodeling of the premature lung.

Although these studies detected pathways that led to aberration in the premature lung, neither study contributed to the effect on hemodynamics or long-term outcomes.  These two studies of varying gestational ages suggest that gestational age, low birth weight, multiple births, and mechanical ventilation affect ventilator acquired lung injury prior to term gestation.

Hemodynamics

Hemodynamic Effects of Mechanical Ventilation

Recent articles on the effects of mechanical ventilation on hemodynamic function are lacking.  Song, et al. (2014) conducted a validation study of 108-paired measurements and found the percentage error measured during various modes of ventilation increased error percentage during the high-frequency oscillator therapy creating so much interference the data was not usable.

This isolated study was difficult to interpret as there is no study evaluating the percentage of error with transthoracic echocardiography during HFO therapy and both technologies have approximately a 30% mean error percentage.

Nursing Assessment of Hemodynamics in Premature Infants

In reviewing the neonatal nursing literature there are no articles that investigate the use of EC.  Currently, hemodynamic monitoring in the nursing literature for premature infants includes heart rate, respiratory rate, saturated pulse oximetry, and noninvasive blood pressure.  There is no stratification by gestational age.  Several authors have contributed to the understanding that these measurements do not directly correlate with changes in stroke volume, cardiac output, and systemic perfusion in premature infants leading to delayed, misleading, and inaccurate measurement (de Boode, 2010; Soleymani, Borzage, & Seri, 2012).

Electrical Cardiography

Current methods available for noninvasively measuring hemodynamic status in premature infants include transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) using ultrasound, and measurements using thoracic electrical bioimpedance using EC via the ICON^® and Aesculon^® from Osypka Medical which is the only impedance device approved for use in premature infants.

Normative data

Poster presentations.  Three poster presentations were located within the existing body of science on EC.  Bozeman (2009) presented a poster on a study looking at fifty stable premature infants including normal cardiac out measurements and inter-rater reliability testing.  Bao Ho (2009) investigated and presented a study of 75 stable premature infants at the Federation of American Societies for Experimental Biology 2009, providing normative data and feasibility results.  Eric Long (2012) studied 105 stable premature infants and provided further acceptable feasibility and reliability data, and normal stroke volume and cardiac output values that were congruent with Bozeman (2009) and Ho (2009) at the Southeastern Surgical Conference.

Published research.  Published peer-reviewed research included a recent article by Hsu et al. (2016).  Hsu and colleagues conducted a pilot study in 280 stable premature infants investigating normative data.  Measurements were taken by gestational age and included cardiac output and index, heart rate, stroke volume, thoracic fluid content, index of conductivity, and systemic vascular resistance.

The posters and recent normative data study evaluated stable premature infants across gestational ages from 23 to 37 weeks gestation.  The reliability and feasibility data of 155 premature infants presented hemodynamic measurements that were congruent with the work from Hsu et al. (2016).  The normative data is robust and includes premature infants of varying gestational ages with consistent findings.  The literature is lacking in premature infants who are unstable or undergoing mechanical ventilation.  Changes in intrathoracic pressure and lung volume whether active or passive affect the cardiovascular system and few studies investigated this phenomenon.

Comparison to Echocardiography

The literature contained several reports of investigations looking at EC for validation of the measurements against the reference equipment currently used which is echocardiography.  The search of the literature yielded eight studies that looked at validation, bias, correlation, and normative data (Table 5).

There were eight studies totaling 1,453 measurements comparing TTE and EC.  Descriptive data was comparable, with no reported racial or gender differences.  Statistical methods for testing limits of agreement included Bland-Altman method, correlation coefficients, and Intra-class correlation.  Correlation coefficients ranged from 0.51 to 0.83.  Pearson’s correlation coefficients ranged from 0.10 to 0.378 (p < .05).  Five authors quantified agreement by comparing bias (Blohm, et al., 2016; Grollmus, et al., 2012; Grollmus & Gonzalez, 2014; Lien, Hsu, Chu, & Chang, 2014; Song, et al., 2016).  Two studies reported high inter- and intra-rater reliability and test-retest reliability which ranged from 0.85 to 0.92 (Blohm, et al., 2016; Boet, Jourdain, Demontoux, & De Luca, 2016).  Mechanical ventilation in one study found the use of a high-frequency oscillator significantly altered mean percentage error (Song, et al., 2016).  Cardiac output measurements ranged from 115-200 cc/kg/minute.  Three studies reported intra-observer and interobserver reliability including the mean, standard deviations, and Lin’s correlation coefficient.

A wide range of premature infants was included in the analysis of the literature.  Six hundred and one premature infants from 23 weeks to 41 weeks gestation and weights ranging from 649 to over 4 kilograms were included in the review.  Several studies concluded that echocardiography is comparable to EC for measurement of hemodynamic functions and may serve as a useful alternative to echocardiography (Blohm, et al., 2016; Cayabyab, Bhombal, & Ebrahimi, 2011; Grollmus, et al., 2012; Grollmus & Gonzalez, 2014; Lien, Hsu, Chu, & Chang, 2014; Song, Rich, Kim, Finer, & Katheria, 2016).  Limitations found include the use of mainly stable premature infants, few infants between 2 and 3 kilograms, and a lack of comparison of effect sizes between studies.  The normative data was predominantly Caucasian (65%) followed by African American (25%) with few minorities.  The majority of the premature infants in the studies were not being mechanically ventilated.

Table 5:  Correlation with Echocardiography
Reference	Year	Reference versus Comparator	Sample	Weight (grams)	Age	Sample	Mechanical ventilation	Bias	Mean % Error	Correlation
Blohm, et al.	2016	TTE / EC	72	800-860	2 days-17 years	285	Yes	< 10%	< 30%	r = 0.10  (p = 0.001)
Boet, et al.	2011	TTE / EC	79	1500-1800	31 weeks	100	Yes		15%	r = .82
Cayabyab,et al.	2011	TTE / EC	48	800 – 2200	23-41 weeks	50	Yes		20%	r = .86
Grollmus, et al.	2012	TTE / EC	24	3330  + 0.5	10 days (3-29)	240	Yes	0.28 (7.6%)	29%	r = 0.14  (p = .01)
Grollmus, et al.	2014	TTE / EC	28	1.618 + 0.346	3.7 + 3.1 weeks	110	No	8.9%	< 10%	r = 0.79
Hsu, et al.	2016	NA	280	800-4420 grams	26.5-41.4 weeks	280	No
Lien et al.	2014	NA	30	929 + 280 grams	27 + 2.0 weeks	280	Yes	< 10%	< 24%	NA
Song, R., et al.	2014	TTE / EC	40	1.072 grams (539-1596)	27 (23-31) weeks	108	Yes  HFO did not correlate	> 10%	< 27%	r = 0.326,  p < 0.005 to r = 0.378, p < .005

Advantages of EC

The literature highlights several advantages revolving around the use of EC.  Benefits of EC included dynamic versus static measurements, the ability to trend the premature infant’s status, noninvasive application, and ease of use (Noori, Drabu, Soleymani, & Seri, 2012; Rauch, et al., 2012).  The published literature contained no reports of adverse events or complications.  Overall percentage error was lower than the clinically acceptable 30% limit (Blohm, et al., 2016).  Several authors found EC accurate, reproducible, and reliable from 23 weeks through 41 weeks gestation.

EC is shown to be a tested noninvasive technology that can be used for the determination of cardiac output monitor in premature infants (Noori, et al., 2012; Rauch, et al., 2012).  EC is reported as a reliable model applied to impedance cardiography to monitor cardiac output and stroke volume noninvasively (Blohm, et al., 2016).  EC is continuous and based on the fact that conductivity of the blood in the aorta changes during the cardiac cycle (Azhibekov, Noori, Soleymani, & Seri, 2013).  Noninvasive cardiac output was reported as a more sensitive indicator of tissue oxygenation and global hemodynamic status that is comparable to other methods (Ballestero, et al., 2011).

Limitations of EC

Several authors found EC more sensitive to external interference than invasive techniques and sited movement artifact as a problem (Blohm, et al., 2016; Boet, Jourdain, & Demontoux, 2016; Song, et al., 2016).  Lien et al. (2014) and Song et al. (2014) identified the need for training in the application of the electrodes and interpretation of the data.  There is a lack of published data in medicine and no published data in the nursing literature on the use of hemodynamic measurements assessed through EC of mechanical ventilation on premature infants.

Summary of the Literature Review

In summary, the neonatal intensive care unit is dependent on advanced technology to support the fragile and unique cardiopulmonary status of the premature infant.  Over 50% of premature infants require mechanical ventilation and advanced forms of life support to survive (Raju, Stevenson, Higgins, & Stark, 2009).  Optimal mechanical ventilation requires a closely monitored balance to support the delicate architecture of the immature brain, lung, and heart.  Mechanical ventilation, although lifesaving is fraught with challenges related to oxygen administration, epigenetic changes in gene expression, altered organogenesis, chronic disease, as well as ventilator acquired lung injury.

Multiple unique anatomical and physiologic variations in the neonatal population restrict approaches to assessment, technology, and care used in pediatrics or adult populations.  Literature exploring the hemodynamic measurements assessed through EC of mechanical ventilated premature infants while controlling for gestational age are limited.  Lower limits of viability present unique challenges in the care of the premature infant.  Survival from the neonatal intensive care unit is no longer the acceptable goal and requires further research to investigate factors that contribute to the short- and long-term morbidities and mortality.

In synthesizing current practice as it affects the research question, anemia is a frequent problem in the neonatal intensive care unit and affects hemodynamic status including systemic blood flow and the filling pressures of the heart.  Strategies to improve erythrocyte endowment are proactive and promising but lack randomization, longitudinal studies, the impact on hemodynamics, and determination of which infants would benefit most.  Surfactant therapy for the treatment of RDS has drastically improved mortality but is confounded by the demographic change of decreasing gestational ages.  There are inconsistent findings regarding the hemodynamic effects of this intervention.  Gentler forms of mechanical ventilation are congruent and do not affect hemodynamic measurements in stable premature infants but require further investigation in unstable premature infants.  These confounding factors must be taken into consideration to explore the specific hypothesis in this study.

Current neonatal nursing practice includes measurements of heart rate, noninvasive blood pressure, and nursing assessment used for the last three decades.  A thesis by Willem-Pieter de Boode (2010) discusses the inaccuracy of these current methods in assessing hemodynamic measurements in newborn infants.  The literature indicates that a low cardiac output contributes to increased risk of cerebral injury, neurodevelopmental deficits, and damage to the immature organs.  Every week of gestation has unique genetically coded stages of development that are exquisitely sensitive to the extrauterine environment.  Neonatal nurses can no longer rely on technology that provides indirect, late, or inaccurate measures of hemodynamic in this fragile population.

Premature infants requiring mechanical ventilation continue to constitute a high-risk population in the neonatal intensive care unit.  Premature infants requiring mechanical ventilation are at increased risk of short- and long-term morbidity and mortality with the very low birth weight and lower gestational aged premature infants suffering the highest morbidity and mortality rates.  The rate of hospitalizations and lingering developmental delays after discharge from the neonatal intensive care unit continues to be of concern for premature survivors.

Theoretically, EC is a technology that uses the alignment of erythrocytes to measure indices of hemodynamics.  Trials have investigated safety, side effects, and gestational age ranges in initially small and now larger samples of premature infants.  Larger trials have confirmed feasibility, reproducibility, and comparison to Echocardiograph.  Research on normative data is increasing, and EC has entered the operative theater.

The measurement of hemodynamics is of considerable importance in the management of the fragile and often critically ill premature infant under intensive care.  There is sparse research demonstrating that EC improves the outcomes of premature infants.  No prior study exists to examine the specific question in the study.  Neonatal nurses require accurate, continuous, dynamic, and individualized measurements to institute the most appropriate nursing interventions based on the integration and synthesis of all currently available data of the exquisitely vulnerable premature infant.  Mechanical ventilation poses many risks to the premature infant and demands accurate, appropriate monitoring and measurements to minimize short- and long-term sequela.  Only by understanding the effect of the environment, therapies, and impact of nursing care can we begin to improve short- and long-term outcomes and demonstrate the necessity and most beneficial application of new technology into the praxis of nursing.

 

 

 

 

 

 

 

Chapter III

Methodology

Prematurity represents a significant financial burden on the family and health care system.  Premature infants are at greater risk of developing negative short-and long-term outcomes including chronic lung disease, decreased blood flow to major organs, and death (Natarajan & Shankaran, 2016; Patel, 2016).  Hemodynamic measurements such as stroke volume and cardiac output are important indicators of cardiovascular function.  The overall goal of mechanical ventilation in premature infants remains the provision of an adequate supply of oxygen in relation to oxygen consumption.  To achieve this goal, a clear understanding of the complex and interrelated interactions of the cardiopulmonary system is essential to optimize care and minimize the risk of complications.  Electrical cardiometry measures cardiovascular performance providing measures of hemodynamics related to changes in erythrocyte orientation.  Mechanical ventilation alters intrathoracic pressure and is transmitted to the immature lungs and myocardium which in turn alters the impedance of blood flowing through the chest.   Achieving the optimal balance between oxygenation and cardiopulmonary function requires knowledge of the impact of the mechanical ventilator on this relationship.  Chapter III focuses on the methodology including a brief overview of the Chapter II literature review, design including threats to internal and external validity, ethical considerations, sampling plan, instrumentation, procedure, and data analysis techniques.

Literature Review

Published research on the use of electrical cardiometry in premature infants reflects the investigation of normative data, correlation of measurements with weight and length, and comparison with echocardiography (Blohm, et al., 2016; Hsu, et al., 2016; Torigoe, et al., 2015).  Lacking in the current literature is research in premature infants on the use of electrical cardiometry, assessment of hemodynamics, and impact of mechanical ventilation on hemodynamic parameters creating a need for more beginning research exploring potential relationships.

Introduction and Statement of Research Problem

The neonatal intensive care unit presents the clinician with an array of rapidly developing premature infants creating many challenges secondary to the immaturity of multiple organs and their fragile susceptibility to short-and long-term complications.  Remarkable progress has been made to support the care of this population, increasing survival, but challenges remain for the bedside clinician.  Evidence on the impact of mechanical ventilation on hemodynamics assessed through electrical cardiometry, in the premature infant, does not exist.  Investigating hemodynamics evaluated through electrical cardiometry under the effects of changes in intra-thoracic pressure from mechanical ventilation requires further evidence on the impact on morbidity and mortality.  Currently, echocardiography is used to assess hemodynamic performance, but these results do not aid the bedside clinician as this technology provides an isolated measurement at one point in time.  Although current treatment of cardiopulmonary failure has increased survival, the impact on morbidities and mortality remain elusive.

Theory-Practice Linkage

Electrical cardiometry represents the scientific area under investigation.  Based on the theoretical Model of Electrical Velocimetry, thoracic electrical bioimpedance captures hemodynamic parameters including measurements of cardiopulmonary function through changes in the orientation of erythrocytes during different phases of the cardiac cycle (Bernstein & Osypka, 2001; Osypka & Bernstein, 1999).  Electrical cardiometry uses four surface electrodes to noninvasively capture hemodynamic measurements through waveforms derived from thoracic electrical bioimpedance.  Through the surface electrodes, the peak impedance derivative of the premature infants blood flow through the aorta occurs simultaneously with the peak acceleration of blood.  In a stable state of volume, the orientation of the erythrocytes causes resistance changes in velocity and acceleration and provides evidence of the effect of mechanical ventilation on hemodynamic measurements.

Despite all we know, no research exists on the hemodynamic effects, assessed through electrical cardiometry, of mechanical ventilation on the premature infants.  The aim of this study is to prospectively examine hemodynamic measurements using electrical cardiometry in a cohort of premature infants receiving mechanical ventilation.  This study hypothesizes that the application of electrical cardiometry for hemodynamic measurement during mechanical ventilation will establish the strength and direction of relationships between and among variables in a natural setting as a precursor to further exploratory and interventional research investigating causation.  This correlational research will determine if a relationship exists between variables supported by a theoretical model and if more rigorous research is warranted.

Hypotheses

The study includes three aims and hypotheses to show the directionality of the aim.

Aim 1.  To evaluate the relationship between the hemodynamic effects, assessed through electrical cardiometry, of mechanical ventilation in premature infants controlling for gestational on the probability of death from respiratory failure.

H_0.  Electrical cardiometry will demonstrate no relationships between the hemodynamic measurements of mechanical ventilation and the probability of death from respiratory failure in a cohort of premature infants.

Aim 2.  To evaluate the correlation between gestational age and indicators of hemodynamic decompensation in mechanical ventilated premature infants using electrical cardiometry.

H₀.  Electrical cardiometry derived hemodynamic measurements of mechanically ventilated premature infants will not demonstrate a correlation between gestational age and indicators of hemodynamic decompensation in contractility, flow, fluid, and/or resistance.

Aim 3.  To investigate the ability of electrical cardiometry measurements of hemodynamic function in premature infants receiving mechanical ventilation to predict outcomes.

H₀.  Electrical cardiometry assessed hemodynamic function of mechanically ventilated premature infants will demonstrate no predictive value for oxygen dependence at 36 weeks postmenstrual age.

Literature Search Strategy

An initial comprehensive and systematic computer-based search strategy was conducted to investigate the current state of the literature (see pages 34 through 36 of Chapter 2).  The initial selection of articles for this review took place October 22, 2016, guided by the question: “What are the hemodynamic effects assessed through electrical cardiometry of mechanical ventilation on premature infants in a Level III neonatal intensive care unit?”  Databases searched included Cumulative Index to Nursing and Allied Health (CINAHL), EBSCO Federated Search, Web of Science, Scopus, ProQuest, dissertations and thesis through the Catholic University of America, and hand searched abstracts from the Society of Pediatric Research.  The time frame used included articles published within the last the last five years and was expanded to ten years secondary to limited findings.  A combination of controlled vocabulary, words, and phrases was included in the search strategy (see Chapter 2 pages 33 through 36).  Key words included a systematic exploration and refinement of neonate, cardiac output, hemodynamic, electrical cardiometry, and terms reflective of current neonatal practice (see Chapter 2 pages 34 and 35).   MeSH terms and filters were outlined in Chapter 2.  Limitations to the search included available research on the use of electrical cardiometry within the identified neonatal population.  The literature continues to be scanned monthly for recently published research.

The major limitation found in the literature search is a lack of empirical evidence in the neonatal population.

Table Summarizing Prior Studies

Several tables were created to summarize the relevant current literature and are located in Chapter 2 (see Chapter 2, page 41, Table 3 Umbilical Cord Literature and Hemodynamic Effect, pages 43 and 44, Table 4 Surfactant Administration and pages 52 through 54, Table 5 Echocardiographic Hemodynamic and PDA Changes).  The following Table 1 titled a Correlation of Electrical Velocimetry/Cardiometry with Echocardiography summarizes prior studies relevant to the proposed research describing normative data, group comparison and correlation with the current standard of care for hemodynamic assessment in premature infants which is echocardiography.

Table 1    Correlation of Electrical Velocimetry/Cardiometry with Echocardiography
Citation	Year	Study Focus	Research Design	Sample Criteria, Sample Size, & Sampling Approach	Outcomes	Comments
Blohm, et al.	2016	Measurement accuracy between echocardiography and electrical cardiometry with anatomical shunts in the neonatal population	Prospective observational cohort study	Healthy preterm and term neonates spontaneously breathing without respiratory support, inotrophic medications, and/or cardiac anomalies  99 infants Convenience sample	Feasible  Physiological shunts in premature infants affect accuracy of left ventricular measurements resulting in an increased bias and overestimation left ventricular stroke volume	Weaknesses: Single center, convenience sample without predefined sample size  Strengths: IRB approved, feasibility, beginning descriptive work with method comparison Bland-Altman analysis Electrical cardiometry equipment was the Aesculon
Boet, et al.	2016	Comparing correlation and variability between echocardiography and electrical cardiometry in hemodynamically stable premature infants	Prospective correlational cohort study	Stable premature infants without cardiovascular compromise  79 premature infants Convenience sample	Good correlation to echocardiography with stroke volume and cardiac output	Weaknesses: Single center, lack of justified sample size, convenience sample  Strengths: Blinded, IRB approved, method comparison Bland-Altman analysis
Grollmus, et al.	2012	To compare electrical velocimetry with echocardiography	Prospective  observational cohort study	Stable preterm infants after heart surgery  24 newborns after switch procedure Convenience sample	Acceptable bias, precision and limits of agreement between echocardiography and thoracic electrical bioimpedance	Weaknesses: Small sample size unjustified, convenience sample  Strengths:  IRB approved, method comparison Bland-Altman analysis
Grollmus, et al.	2014	To compare electrical velocimetry with echocardiography	Prospective  observational  cohort study	Cardiac output measurement in low and very low birth weight infants  228 premature infants Convenience sample	Electrical velocimetry is safe, easy, and comparable to echocardiography	Strengths:  Multi-center study,  IRB approved, method comparison Bland-Altman analysis, precision calculation of the coefficient of variation
Hsu, et al.	2016	To investigate hemodynamic references data for age, weight, and hemodynamic measurements	Prospective observational  cohort study	Hemodynamically stable premature infants excluding cardiac defects  280 premature infants Convenience sample	Hemodynamic normative data provided  Cardiac output correlates closely with gestational age	Weakness:  Convenience sample  Strengths:  Multi-center, IRB approved, sample size Regression analysis
Hsu, et al.	2017	To investigate agreement between electrical cardiometry and echocardiography	Retrospective correlational cohort  study	Hemodynamically stable premature infants  36 preterm infants Infants with left to right shunt were enrolled	Electrical cardiometry and echocardiography acceptable agreement in measuring hemodynamics in premature infants with a patent ductus arteriosus however infants with high cardiac output or ventilated by high frequency ventilation require caution	Secondary data  Caution with high frequency ventilation Strengths:  Multi-center, method comparison Bland-Altman analysis Weaknesses:  Small sample size, lack of predefined sample size
Lien, et al.	2014	To examine hemodynamic effects during patent ductus arteriosus ligation in very low birth weight infants	Prospective observational cohort study	Inclusion criteria were premature infants undergoing surgical ligation of a hemodynamically significant the patent ductus arteriosus with measurements one hour prior to surgery and 48 hours after surgery  30 premature infants during patent ductus arteriosus ligation Convenience sample	Measurements included stroke volume, heart rate, cardiac output and index and systemic vascular resistance  Description of hemodynamic baseline measurements and respective changes using electrical cardiography and echocardiography	Weaknesses: Small sample size of surgical infants with large patent ductus arteriosus,  Single center, convenience sample Strengths:  Electrical cardiometry equipment was the Aesculon Statistical analysis included repeated measures with Bonferroni correction, multiple regression and specific time point of each subgroup were compared using independent t-test
Song, et al.	2014	To investigate utility and correlation with echocardiography	Prospective correlational  cohort study	Premature infants less  than 32 weeks gestation without congenital heart disease (except small VSD), inability to tolerate adhesive skin leads, and the presence of major congenital anomalies  40 premature infants Data was prospective and part of a larger trial investigating cord milking for erythrocyte endowment	Electrical cardiometry correlates with both right and left ventricular outflow with limitations at low output states and with high frequency ventilation	Weaknesses:  Single center, small sample size, convenience sample  Paired t-test for parametric data Pearson coefficient correlation Method comparison Bland-Altman plot

Analysis of the Literature and Limitations

There is limited published research on the use of electrical cardiometry in the neonatal population.  The majority of research has compared electrical cardiometry with the current standard of transthoracic echocardiography.   Prior validation studies have found acceptable correlation between electrical cardiometry and transthoracic echocardiography (Blohm, et al., 2016; Boet, Jourdain, Demontoux, & De Luca, 2016; Grollmus, et al., 2012; Grollmus & Gonzalez, 2014; Hsu, et al., 2017; Song, et al., 2014).  Clinical evaluations have additionally shown repeatability and the ability to trend hemodynamic measurements over time (Grollmus & Gonzalez, 2014; Hsu, et al., 2016).   Prior designs have investigated observational data to describe normative data and the interchangeability of electrical cardiometry with echocardiography for estimating hemodynamic measures.   From the literature review, Bland-Altman analysis was most often used for method comparison (n=536) (Blohm, et al., 2016; Boet, Jourdain, Demontoux, & De Luca, 2016; Grollmus, et al., 2012; Grollmus & Gonzalez, 2014; Hsu, et al., 2017; Song, et al., 2014).  Normative data for electrical cardiometry has been described by several authors comparing gestational age, weight, and body surface area (Hsu, et al., 2016; Song, et al., 2014).  The research of prior scientists have provided the platform for further investigation and provided the knowledge necessary to further explore and contribute to beginning research necessary to ascertain if more rigorous research is warranted.

Threats to Internal Validity.  Because this study uses physiologic measures internal validity is concerned with accuracy, precision, and error (Gray, Grove, & Sutherland, 2017, p. 384).  Accuracy and precision of electrical cardiometry with echocardiography has been established in multiple studies.  Two studies found that physiological shunts affected the measurement accuracy of electrical cardiometry.  Infants with underlying cardiac anomalies and a clinically significant patent ductus arteriosus will be excluded (Blohm, et al., 2016; Song, et al., 2014).  Two studies identified a lack of accuracy between electrical cardiometry and echocardiography when infants were ventilated with high-frequency ventilation which creates movement distortion in the waveform affecting measurements.  High frequency oscillating ventilation will be included as an exclusion criterion (Hsu, et al., 2017; Song, et al., 2014).  Threats will be minimized with a detailed protocol, inter- and intra-rater reliability, automated data transfer, signal quality indicators, and waveform analysis.  Precision will be supported by comparing heart rate measurements between the bedside monitor and the impedance device.  Additionally, the neonatal intensive care unit is familiar with the technology, and the primary investigator has greater than ten years prior experience in use of the technology and thirty years of practice in neonatal nursing.

Threats to External Validity.  The results of the study may be generalizable to 23 to 37 weeks gestation infants who are mechanically ventilated in a Southeastern metropolitan Level III neonatal intensive care unit that fall within the inclusion and exclusion criteria.  To reduce and minimize threats to external validity a detailed and rigorous procedure will be outlined including an adequate sample size and the assistance of a biostatistician.

Research Design

The architecture of the study is quantitative and will use a prospective correlational design.  Primary data gathered is based on the theoretical Model of Electrical Velocimetry (Bernstein & Osypka, 2001; Osypka & Bernstein, 1999).  The correlational design allows numerical measurement of the strength of relationships between and among variables to discover if there is a change and if so, whether that change in one or more variables happens when another variable decreases or increases.  Additionally, the study investigates the influence of demographic data on these relationships and controls for gestational age because of varying degrees of organ maturity and development.

Variables

The independent variables in this study include five continuous independent variables of contractility, flow, fluids, resistance, and mechanical ventilation (See Appendix F Statistical Analysis Plan).  Contractility, flow, fluids and resistance are obtained using electrical cardiometry.  The three categorical dependent variables are oxygen dependence at 36 weeks gestation, death from respiratory failure, and hemodynamic decompensation measured as one standard deviations above the mean.  Covariates include gestational age, race, and demographic variables.  Demographic variables captured from the electronic medical record include APGAR score, gestational age, gender, birth length, race, and birth weight.  The purpose of this design is to examine the relationships that occur between and among variables and establish the existence of relationships.

The investigation does not attempt to manipulate or exert control over the mechanically ventilated infants studied.  In this study, the data consists of observations and measurements with no randomization, treatment, control group or direct manipulation of the variables.  The study collects a number of hemodynamic measurements using electrical cardiometry in premature infants receiving mechanical ventilation and examines hemodynamic measurements and the relationships between and among the measurements.  No published research has investigated the use of electrical cardiometry, assessment of hemodynamics, and impact of mechanical ventilation on hemodynamic parameters creating a need for exploratory and beginning research.

Inclusion and Exclusion Criteria

Inclusion criteria will be premature infants admitted to a Level III neonatal intensive care unit less than 37 weeks gestation.  Exclusion criteria will be conditions that cause potential sensor interference (anasarca, congenital anomalies, or hyperbilirubinemia), potential for altered perfusion (APGAR score less than 3 at 5 minutes, congenital heart disease, necrotizing enterocolitis, sepsis), and impedance cardiogram distortion (large patent ductus arteriosus, high frequency oscillating and Jet ventilation).  A convenience sampling method will be used for this study for logistic reasons.

Protection of Human Subjects

Institutional Review Board Approval

Multiple conditions are met to protect the diminished autonomy of the premature infant.  The study will be Institutional Review Board approved and deemed a scientifically appropriate study.  Approval will be obtained through the Institutional Review Boards at the Catholic University of America and the Medical Center of Central Georgia for the protection of human subjects.  The nurse investigator will have completed and placed on file a copy of the Collaborative Institutional Training Initiative (CITI) training in human subject’s research protections, and the responsible conduct of research designated by both institutions.

Risk

The investigation presents minimal risks to the subjects.  The sensors used will be water soluble and only placed on intact skin.

Parental Consent

Fully informed parental consent will be obtained.  If there are two parents, they both will be fully informed and uncoerced to provide informed consent.  A separate Health Insurance Portability and Accountability (HIPPA) consent form will be obtained to gather demographic data from the medical record using only the random subject number as an identifier.  Both parents will be provided a copy of the consent forms including the phone number to the primary investigator.

Maternal Age.  In accordance with institutional policy at the Medical Center of Central Georgia mothers younger than 18 years old will require additional parental consent.  Infants of mothers younger than 13 years of age will be excluded from the investigation per hospital policy.

Confidentiality

All data collected will be secured to protect confidentially.  The right to confidentiality of all data collected will be protected with the use of a randomized research subject number and removal of the 18 types of identifiers (Gray, Groves, & Sutherland, 2017, p. 169).  A master list of subject names and their code numbers will be kept in a locked office in a password protected computer research database.  All documents will use the subject number identifier, and other neonatal practitioners will not have access the data gathered.  Confidentiality of the subject’s information will be protected during the data analysis process and the names and code number list destroyed upon completion of the study.

Data Storage and Data Security

Confidentiality of the subjects’ information will be assured during the data collection, storage, and analysis process.  The ICON^® will store all data collected at the bedside with a random subject identification number in a password protected computer.  Daily subject data will be uploaded to the blinded biostatistician using an asymmetric encryption algorithm with two keys to share and assure confidentiality, integrity, and authenticity of electronic data communications.  Stored data will be protected from physical damage, loss or theft through limited access to the data, password authorization, and daily uploads to the biostatistician.  To protect the computer system up-to-date firewall, anti-virus protection and software will be used on the research computer.  To maintain the integrity of the stored research data, the off-site biostatistician will be used as backup storage during the research project.

Sampling Plan

A nonprobability quota sampling method will be used in this prospective correlational design.  Quota sampling ensures that an even distribution of infants in each weight range is represented in a population that may not be represented in a convenience sample.  This sampling method provides an adequate number of premature infants in each stratum for the statistical analyses.  This approach is justified because the quota sampling will stratify for a more homogenous sample in addition to reducing the potential for error from extraneous variables as extremely lower birth weight premature infants have higher mortality rates (See Appendix A Sampling Strata).

Recruitment Strategy

Recruitment will be face-to-face in the neonatal intensive care unit and enrollment is anticipated to open June 2018 after Institutional Review Board approvals.  Fully informed consent will be obtained from the parents in addition to Health Care Privacy consent for access to the medical records.  Monthly admissions of premature infants requiring mechanical ventilation for respiratory distress syndrome average approximately 100 premature infants per month.

Setting

The setting for this single center investigation is a private, not-for-profit 650 bed academic teaching hospital with a Level I Trauma center and dedicated children’s hospital.  The facility is located in the southeastern United States with a population of 160,000 located within a state with a population of 3.7 million (Census Viewer, 2017; United States Census Bureau, 2017).  The neonatal intensive care unit is a Level III perinatal center that routinely cares for mechanically ventilated premature infants.  English speaking parents of mechanically ventilated premature infants greater than 72 hours of age will be recruited.  Mothers younger than 18 years old will require additional parental consent per hospital policy.  Mothers younger than 13 years of age will be excluded per hospital policy.  The 72 hour enrollment window reflects the time needed for the intrauterine shunt called the patent ductus arteriosus to close and the first echocardiogram to be completed to assess for congenital heart disease and a significant right to left shunt which potentially could distort the impedance cardiogram.  To show respect and appreciation to the staff of the neonatal intensive care unit and parents whose infants were enrolled those wishing to know will be provided the results and a ‘thank you’ from the researcher for helping to build the science necessary to care for premature infants.

Sample Size

An a priori sample size estimation was calculated for logistic regression models using a 90% power, significance criterion of α = 0.05, for five predictor variables and two covariates using G*Power 3.1^TM (See Appendix B G*Power Analysis).  The calculated sample size was 170 and added to the sample size is an attrition rate of 20% making the total sample size 204 subjects.

Instrumentation

Electrical cardiometry derived hemodynamic measurements will be obtained using four ICON^® workstations (Cardiotronic-Osypka Medical, La Jolla, CA/Berlin, Germany).  The sensors used with the ICON^® to capture the measurements will be the iSense-neonatal three centimeter (cm) skin sensors individually packed (Cardiotronic Sensors^TM for Electrical Cardiometry^TM).  For weight and length, a calibrated Tanita BD-815U neonatal scale with accuracy measurements within + 4 grams for weight and a calibrated Seca 207 measuring rod for length will also be used.  Biomedical engineering will certify all equipment and recheck in the first week of each month during data collection.  The equipment is for research purposes only for use by the data collectors.

Accuracy

To assure accuracy, the ICONs^® will be checked by biomedical engineering prior to enrollment.  A detailed plan of data collection will ensure each data collector follows the same protocol to standardize the process.  All data collectors have a minimum of two years’ experience using the ICON^® and have a competency form signed and dated in the research cabinet (see Appendix D Competency Form).  The ICON^® runs through a self-check prior to application to the research subject.  Once connected to the research subject, the ICON^® heart rate is compared to the heart rate on the bedside monitor and should be the same + 4 heartbeats.  The ICON^® signal quality should be at 100%, and impedance waveforms will be assessed for clarity.

Procedure

The research procedure is as follows in accordance with the inclusion and exclusion criteria:

1. Informed consent and the HIPPA addendum will be obtained from both parents as applicable; both parents must be in agreement to proceed.
2. Once enrolled the premature infant will be provided a randomized subject identification number and entered into the research log book maintained in a locked cabinet in a password protected office and password protected research computer.  No infant identifiers will be used in the study.  The randomizer is part of the hospital research platform available to all researchers.
3. Data collectors will do a one minute scrub upon entrance to the neonatal intensive care unit at the scrub stations.  Bedside gloves will be worn while working with the premature infant.  Hands will be washed with bedside antiseptic per unit protocol before and after touching the premature infant.  For infants in isolation, unit protocol will be followed.
4. Demographic data to compare the population and investigate confounders will be gathered from the electronic medical record, verified with the parents, where indicated (race) and entered into the research computer.  Data derived from the electronic medical record includes the APGAR score, gestational age of the premature infant, gender, hemoglobin level, birth length, race and birth weight.
5. Anthropometric measurements will be obtained by nurse researchers the day of the study using a balanced and warmed Tanita BD-815U neonatal scale for weight and a Seca 207 measuring rod for length.
6. The ICON^® workstation, kept in the research office in a clean bag, and will be brought to the bedside and plugged into a three prong bedside plug.  The ICON^® is preprogrammed to collect data every 5 minutes for the identified study variables.
7. A monthly research calendar posted in the neonatal intensive care unit and on the research computer will identify which two data collectors are covering the neonatal intensive care unit on the day shift and night shift.  The research team will provide 24 hour coverage through data collection.
8. The data collector will work with the bedside nurse to coordinate working with the premature infant around established touch times and care.
9. To apply the sensors to the infant’s skin, the skin will be gently wiped with a four-by-four gauze sponge.  New and in date iSense-neonatal skin sensors will be placed on intact dry skin of the premature infant.  The array of four sensors measures voltage across the middle two sensors; one will be placed at the bifurcation of the neck and another at the fourth intercostal space to bracket the ascending aorta.  The other two outer current sensors will be placed 2.0 cm apart on the mid-frontal scalp and left mid-thigh.
10. The nurse researcher will press menu on the ICON^® and connect the research computer to the system for computerized uploading of the measurements verifying the signal quality, heart rate measurements with the bedside monitor, and visualization of the impedance waveform.  The Institutional Review Board study number and randomized subject number will be entered into the ICON^® workstation including the anthropomorphic data.
11. Data will be collected for eight continuous hours.  Subject event codes loaded into the ICON^® will be used to describe the infant’s environment, state, and any procedures that may influence the measurements secondary to movement, pain, or procedural artifact (see Appendix C ICON^® Comments Markers).
12. Sensors that become dislodged will be replaced with a new sensor by data collectors.
13. Once the data collection on the patient is completed, the data will be encrypted and uploaded to the biostatistician and a second back-up copy kept on the research computer.
14. The data collector will coordinate with the bedside nurse to remove the electrodes with cotton balls and warmed water inspecting the skin for any redness or irritation which will be reported to the primary investigator and medical director.
15. The ICON^® workstation will be cleaned at the bedside using hospital approved Cavicide disinfectant wipes and placed in a clean equipment bag and returned to the research office.  If for any reason the equipment cannot be cleaned immediately it will be placed in the dirty equipment room.
16. The data collector will log the date and time of the study and status including any issues, problems or concerns.
17. Parents will be notified of the data collection completion and thanked by the data collector.
18. Video updates will be provided to the neonatal intensive care unit and nursing research council by the primary investigator monthly through the research computer program.
19. The primary investigator will work with the offsite biostatistician to support data analysis.  Once the data are analyzed the research team will meet to discuss findings and further action.

Precision

Quality control testing of the ICON^® is determined by the manufacturer internal quality control and is automatically done within the equipment prior to the connection with each new patient.  The Intra-Class Correlation for inter-rater reliability (IRR), a standard technique that tests raters as a random sample of all possible raters, will be utilized.  IRR will demonstrate consistency across researchers collecting measurements on the premature infants.  Reliability of the measurements between data collectors will be critical to the focus of the paper.  Low IRR could increase the risk of Type-II errors.  The analysis of IRR will be used to determine the variance between operators due to inaccuracy or error.

Error in Physiologic Measurements

There are potential sources of error including user, research participants, and equipment (Gray, Grove, & Sutherland, 2017).  Data collectors will demonstrate competency and have a form on file in the neonatal intensive care unit research office (see Appendix D Competency Form).  To capture the influence of the environment coded variables will be used and entered by the data collector.  Each data collector will remain at the bedside for the eight-hour study.  A second researcher will be paged to cover time away from the research study and verify demographic data entry.  To remove transcription errors, the hemodynamic measurement data are automatically uploaded into the ICON^® and research computer.  Signal quality is included in the upload for data analysis including EC waveforms.  The biostatistician has worked with the equipment for over ten years and is competent to analyze the impedance measurements.  A doctorally prepared biomedical engineer will support questions related to waveform analysis, interference, or artifact.  Signal quality will be four bars or 100 percent to represent valid data.

Data Analysis Plan

IBM SPSS Statistics for Windows, version 24 (IBM Corp., Armonk, N.Y.,USA) software program will be employed to conduct the statistical analysis.  The data analysis plan and methods are supported by a biostatistician.

Data Entry

Data dictionary.  A data dictionary will be created in SPSS 24 and will include the codes assigned to each variable including the variables name, measurement, label, type of variable and normative data ranges if applicable (see Appendix E Data Dictionary).

Entering the data.  Subject identification numbers will be entered with the demographic data (APGAR score at one and five minutes, gestational age, gender, hemoglobin level, length, race, surfactant administration with frequency and weight).

Cleaning the data.  The data will be reviewed in the database to detect, correct, or remove incorrect data.  The data collector placing the infant on the monitor will enter the data and the second data collector for the shift will review the data.  The next day the data will be reviewed by the primary investigator.

Examining the data.  The data will be examined by the primary investigator and data collector by examining box plots, histograms, and frequencies to detect inappropriate or invalid responses.

Data Analysis Techniques

For this quantitative study with categorical dependent variables, the data analysis plan will be conducted in three phases.  The analysis will begin with descriptive statistics via univariate analysis.  Second, the relationships between the dependent variables with the independent variables and covariate variables will be examined using bivariate analysis.  Independent and covariate variables associated with the dependent variables as statistically significant (p<0.05) will be entered into the final multivariate model.

The Intra-Class Correlation for IRR, a standard technique that treats raters as a random sample of all possible raters, will be utilized for evaluation.  IRR will be used to demonstrate consistency across data collectors.  The IRR statistic was chosen as the best fit for the goal of the study.  IRR was necessary to provide agreement between researchers collecting the data to support the study design and hypothesis.  Reliability of the measurements between data collectors will be critical to the focus of the paper.  Low IRR could increase the risk of Type-II errors.  The investigation and analysis of IRR will be used to determine the variance between operators due to inaccuracy or error.  The statistics used to compute IRR will be the Intra-Class Correlation (ICC) for IRR.  A subset of five subjects will be used to rate multiple data collectors in a two-way model.  Good IRR will be characterized by consistency in the data based on averages of ratings captured by multiple researchers.  ICC incorporates magnitudes of disagreement whereby a larger magnitude of disagreement results in smaller ICCs.  Higher ICCs reflect greater IRR with an estimate of 1 indicating perfect agreement and 0 indicating random agreement.  The ICC value will be set between 0.75 and 1.0.

Inferential statistics will be used to test the research hypotheses.  Descriptive statistics will consist of measures of central tendency and frequencies to describe the sample.  Bivariate correlations will be used to determine which predictors to include in the models.  Separate logistic regression models will be fit to each independent variable against each dependent variable.

With five independent variables, and three dependent variables there will be 15 logistic equations and 15 odds ratios.  These odds ratios are the crude odds ratios and will address each of the three aims on the primary level in that they will tell if each independent variable is related to each dependent variable.  The covariates will be divided into primary and other with gestational age and race as primary cofounders and the remaining as other.  The covariates will be refit into each of the 15 logistic regression equations with the two primary covariates added into the equation (3 models).  This analysis will show if gestational age and race is a moderating effect or a confounding effect by comparing the results.  For the exploratory portion, the remaining covariates will be added to each of the 15 logistic regression equations to see if there are any additional moderating or confounding effects.  The data will be presented as both the raw and Hochberg adjusted p-values offering the reader the ability to make their own judgment.

Equation for Analysis

The statistical model to be analyzed is:

Ŷ(p/1-p) = β₀ + β₁x₁ + β₂x₂ + β₃x₃ + β₄x₄ + β₄x_{5 + error}

Ŷ = dependent variable, categorical

p = probability

Β₀ = intercept

x₁ = contractility, continuous

x₂ = flow, continuous

x₃ = fluids, continuous

x₄ = resistance, continuous

x₅ = other, continuous

Results

Results from the analysis will be presented in tables and figures.

Summary

Prematurity has an enormous impact on the premature infant, family, and society.  Prematurity has high prevalence and significantly contributes to mortality and morbidity.  This investigation will explore hemodynamics assessed through electrical cardiometry of mechanical ventilation in premature infants.  Relationships between the independent and dependent variables will contribute to empirical evidence to better understand these relationships and potentially assist clinicians in determining evidence based guidelines to optimize outcomes.  Additionally this research facilitates understanding of the hemodynamic effects of premature infants receiving mechanical ventilation.  Chapter III has provided a description of the theoretical linkage to the study, design, setting, sample plan, ethical considerations, instrumentation and the general procedure to be used during data collection.  This study aims to unveil the complex and interrelated hemodynamic effects of mechanical ventilation on the premature infant through electrical cardiometry.  A total of 207 premature infants will be used to for this investigation to further the research in this vulnerable population.