Article Text

Assessment of parental decision-making in neonatal cardiac research: a pilot study
  1. Aruna T Nathan1,
  2. K Sarah Hoehn2,
  3. Richard F Ittenbach3,
  4. J William Gaynor4,5,
  5. Susan Nicolson1,4,
  6. Gil Wernovsky4,6,
  7. Robert M Nelson7
  1. 1Department of Anesthesiology and Critical Care Medicine, University of Pennsylvania, Pennsylvania, USA
  2. 2Department of Pediatrics, St Christopher's Hospital for Children, Philadelphia, Pennsylvania, USA
  3. 3Department of Pediatrics, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, USA
  4. 4The Cardiac Center, The Children's Hospital of Philadelphia, Pennsylvania, USA
  5. 5Department of Surgery, The Children's Hospital of Philadelphia and University of Pennsylvania School of Medicine, Pennsylvania, USA
  6. 6Department of Pediatrics, The Children's Hospital of Philadelphia and University of Pennsylvania School of Medicine, Pennsylvania, USA
  7. 7Office of Pediatric Therapeutics, Office of the Commissioner, Food and Drug Administration
  1. Correspondence to Dr Aruna T Nathan, Department of Anesthesiology and Critical Care Medicine, The Children's Hospital of Philadelphia, 12 NW 40, 3400 Civic Center Blvd, Philadelphia, PA, USA; nathan{at}


Objective To assess parental permission for a neonate's research participation using the MacArthur competence assessment tool for clinical research (MacCAT-CR), specifically testing the components of understanding, appreciation, reasoning and choice.

Study Design Quantitative interviews using study-specific MacCAT-CR tools.

Hypothesis Parents of critically ill newborns would produce comparable MacCAT-CR scores to healthy adult controls despite the emotional stress of an infant with critical heart disease or the urgency of surgery. Parents of infants diagnosed prenatally would have higher MacCAT-CR scores than parents of infants diagnosed postnatally. There would be no difference in MacCAT-CR scores between parents with respect to gender or whether they did or did not permit research participation.

Participants Parents of neonates undergoing cardiac surgery who had made decisions about research participation before their neonate's surgery.

Methods The MacCAT-CR.

Results 35 parents (18 mothers; 17 fathers) of 24 neonates completed 55 interviews for one or more of three studies. Total scores: magnetic resonance imaging (mean 36.6, SD 7.71), genetics (mean 38.8, SD 3.44), heart rate variability (mean 37.7, SD 3.30). Parents generally scored higher than published subject populations and were comparable to published control populations with some exceptions.

Conclusions The MacCAT-CR can be used to assess parental permission for neonatal research participation. Despite the stress of a critically ill neonate requiring surgery, parents were able to understand study-specific information and make informed decisions to permit their neonate's participation.

  • informed consent
  • MacArthur competence assessment tool for clinical research
  • newborns and minors
  • research ethics
  • research on special populations
  • understanding

Statistics from

Informed and voluntary consent is, with rare exception, an essential aspect of clinical research.1 2 Parental permission is used as an alternative for neonates to be enrolled in research. Parental permission is understood to include a number of important components: understanding, appreciation, voluntariness and formulation of a choice after considering the risks and consequences of research participation.2 3 In order for parental permission to be deemed valid, important elements should be met—parental comprehension of the study purpose and the risks and benefits to their child and voluntary choice. Voluntary choice is one that is free from undue influence and coercion. These components may be compromised under situations of duress. Also, the perception of adequate time may be important to understanding and processing presented information.4 Concerns are often raised that parents are unable to make an informed decision during periods of great stress.5

In 2000, Mason and Allmark5 performed semistructured interviews with 200 parents who had made decisions about research participation in the newborn period. Overall, 70.5% of parents had problems with one or more of the criteria for valid consent, namely, competence, information, understanding and voluntary choice. More recently, it was reported that 12% of parents who made decisions about participation in a randomised controlled trial did not recall 18 months later making the decision to participate.6 Hoehn et al7 have previously reported that fathers of neonates with critical congenital heart disease diagnosed postnatally had significantly higher levels of anxiety than fathers of neonates diagnosed prenatally. Despite the duress of the situation, parents have often elected to participate in research studies involving their children for societal and potential individual benefits.8 The stress of the diagnosis of a critically ill neonate and the often short time period in which to make decisions about research participation prompted us to explore ways to assess important components of the parental permission process. To date, there is no one particular tool or technique to assess whether parental decisions to allow for research participation are adequately informed.

The MacArthur competence assessment tool for clinical research (MacCAT-CR) has been used to assess the informed consent process in adults with diminished capacity.9 In particular, it has been used with samples of adults with schizophrenia,10–12 depression13 and Alzheimer disease.12–16 Although these are not ideal comparison populations for parents of critically ill children, they are useful in that those studies included adults with diminished capacity and normal controls. To our knowledge, this is the first study in which the MacCAT-CR tool has been used to assess parental permission. Our objective was to use a structured quantitative interview tool designed to evaluate important components of parental permission (ie, understanding, appreciation, reasoning and choice). Our primary hypothesis was that parents of critically ill newborns would have comparable scores on the MacCAT-CR to normative data on adult controls despite the emotional stress of an infant with critical heart disease or the urgency of surgery. We wanted to see if parents of critically ill newborns showed evidence of diminished decision-making capacity, as some have reported,17 especially with regard to understanding the purpose, risks and benefits of the study, and in their ability to make a rational choice under extremely stressful situations.


Study-specific MacCAT-CR interview tools were created from the parental permission documents used with parents of neonates for three different research protocols being conducted in the eligible neonatal population undergoing cardiac surgery. The tools were created by following the specific guidelines of the MacCAT-CR and by taking the language directly from the informed consent documents for each study.9

Study participants consisted of parents of eligible neonates who had made decisions either for or against research participation before their neonate's cardiac surgery. Eligible neonates were those undergoing open heart surgery for critical congenital heart disease on cardiopulmonary bypass. Further eligibility criteria included English-speaking parents, and being offered enrolment in one or more of three research studies (genetics, heart rate variability and MRI studies, described below). Study-specific MacCAT-CR interview tools were used to interview the participating parents. If the parents had been approached for more than one study, they were asked to complete a MacCAT-CR interview about two of those studies. Each interview lasted approximately 30 minutes. Parents were interviewed within 10 days of the neonate's cardiac surgery, and were permitted to refer to the informed consent documents of the study if they so desired. Parents of the same neonate were interviewed separately. Parents were not interviewed consecutively.

All three studies for which the MacCAT-CR was used (described below), as well as the current study, were approved by the institutional review board (IRB) at The Children's Hospital of Philadelphia. Informed consent was obtained verbally for this study, as written documentation was waived by the IRB. A brief summary of the three studies follows.

Genetics study

The purpose of the genetics study was to investigate the genetic aetiology of congenital heart disease. Blood samples were taken from indwelling lines for DNA analysis once, before any form of blood transfusion, and remained available for ongoing analysis. Other research procedures included electrocardiography and/or echocardiography of the parents, genetic examination or digital photographs of the child and or siblings and blood samples from the parents.

HRV study

The purpose of the heart rate variability (HRV) study was to investigate how the autonomic nervous system controls the heart after cardiac surgery performed in infancy. This required placement of a continuous 24 h ECG recording beforeo surgery, before discharge and at 3 months of age, as well as medical record and laboratory data review. This study required additional visits for study purposes alone at 3 months of age.

MRI study

The purpose of the magnetic resonance imaging (MRI) study was to investigate structural brain abnormalities, as well as cerebral blood flow changes in response to inhaled carbon dioxide. This study was the most complex of the three. Additional potential risks to the neonates included an additional 60–90 minutes under general anaesthesia before open heart surgery, two MRI scans, including one with transiently elevated levels of carbon dioxide and another MRI scan at 7–10 days of age under anaesthesia or sedation.18

The MacCAT-CR was initially created for adult patients with psychiatric and geriatric disorders.8 The MacCAT-CR measures the following: understanding (scored 0–26); appreciation (scored 0–6); reasoning (scored 0–8) and choice (scored 0–2), for a total possible score of 42 (table 1). Responses range from poor (0), to fair (1), to good (2). The maximum possible score for each respondent is 42 points. Study-specific MacCAT-CR interview tools were created using the information provided in the consent document for the original studies. All interviews were audiotaped, transcribed, and then corrected for accuracy in the transcription. The scoring was based on an analysis of the transcripts. The interviews were scored by three independent reviewers using study-specific scoring sheets, with initial agreement on 70% of all items. Differences were discussed and resolved, resulting in a final MacCAT-CR score.

Table 1

Elements of the MacCAT-CR*cv9

Data analysis

Data analysis consisted of two components. The first component involved the generation of summary statistics for all relevant demographic and MacCAT-CR variables, including measures of central tendency, variability and association (table 2). For completeness, total and subscale MacCAT-CR scores are presented for each study using both parametric (mean ± SD) and non-parametric (median ± range) measures of central tendency and variability. A series of two-sample t tests were used to test the significance of differences between results obtained in this study with those of other published studies (table 3). Alpha was not adjusted beyond the traditional α=0.05 level due to the descriptive nature of the analyses in this study and the non-overlapping nature of comparisons with other published studies. No comparisons were made between parents who completed two interviews.

Table 2

Measures of central tendency and variability by sample and score

Table 3

Comparison of MacCAT-CR scores of parents of neonates to published MacCAT-CR results

The second component of data analysis evaluated the reliability (internal consistency) of the MacCAT-CR tool itself using the data obtained in this study. Measures of internal consistency, or the ability of items on a scale to work together when evaluating a given trait, were computed using Cronbach's alpha for the entire MacCAT-CR scale for all three studies, and for each of the three subscales individually (understanding, appreciation, reasoning) for the MRI study. The MacCAT-CR scores for the genetics and HRV studies had insufficient variability to compute comparable measures of internal consistency. In addition, measures of association were computed for MacCAT-CR total and subscale scores to estimate the degree of relatedness among scale and subscale scores. To eliminate redundant sampling, the first MacCAT-CR response of each parent was chosen to compute the scale–subscale correlations using Pearson's r correlation coefficients and was evaluated using a two-tailed test of statistical significance.


A total of 35 parents of 18 neonates completed 55 interviews using the MacCAT-CR. Twenty parents were interviewed for two studies and 15 parents for one study alone. None of the parents were interviewed for all three studies. The breakdown by study was 20 for MRI, 16 for HRV and 19 for genetics. The sample of 35 parents consisted of 18 mothers and 17 fathers; 15 of the parents had a neonate who was diagnosed prenatally, whereas 20 parents had a neonate who was diagnosed postnatally. Parents' ages ranged from 15 to 47 years, with an average of 33.4 years (SD 7.4). Self-reported ethnicity was as follows: 31 white Caucasian, two African American, one Asian and one Pacific Islander. The study was limited to English-speaking parents as the MacCAT-CR instruments were in English. The highest level of completed education reported by the parents included: primary (or grade) school (n=3); secondary (or high) school (n=15); college (n=12) and graduate school (n=7). Anatomical cardiac diagnoses of the 18 neonates were as follows: hypoplastic left heart syndrome (n=7); transposition of the great arteries (n=7); total anomalous pulmonary venous return (n=1); coarctation of the aorta (n=1); and tetralogy of Fallot with absent pulmonary valve (n=2). Interviews were conducted between 14 January 2004 and 13 July 2004 on postoperative days 1 to 10. All of the interviews were conducted by two principal investigators (AN, KSH).

Ten parents declined participation in this study. No demographic data were collected on those parents who declined to participate in the study. Of those who initially agreed to participate, three declined to be interviewed after initial demographic data were collected, two parents asked to be interviewed together and data for two parents were lost due to tape-recorder malfunction.

The scores obtained with the MacCAT-CR were as follows (table 2): for the MRI study (n=20), the mean total MacCAT-CR score was 36.65 with a SD of 7.71. The minimum score for the MRI study was 19 and the maximum was 42. For the MRI study, five parents scored below 32 (actual scores 19, 20, 24, 29, 31), which is less than 76% of the items answered correctly (ie, usually considered a passing score). For the genetics study (n=19), the mean total MacCAT-CR score was 38.8 with a SD of 3.45. The minimum was 31 and the maximum was 42. Only one person scored below the 76% mark. For the HRV study (n=16), the mean total MacCAT-CR was 37.7 with a SD of 3.30. The minimum score was 28 and the maximum was 42. Only one parent scored below 32. The breakdown of the MacCAT-CR subscale scores on the three studies (understanding, appreciation and reasoning) is provided in table 2.

When comparing scores for selected subgroups, there were no statistically significant differences in total MacCAT-CR scores between: (1) mothers and fathers; (2) parents who agreed to participate in a research study, and those who refused to participate in a research study; or (3) parents whose neonate was diagnosed prenatally, compared with parents whose neonate was diagnosed postnatally. MacCAT-CR scores between parents with different educational levels or outcomes were not compared, given the limited sample size.

The MacCAT-CR subscale scores (understanding, appreciation and reasoning) of the parents of neonates undergoing cardiac surgery were compared with published values for control and subject populations (table 3). Articles for comparison were selected based on the inclusion of a control group in the published dataset. With the exception of the understanding subscale score of mentally ill prisoners18 and one report from patients with schizophrenia (understanding and appreciation subscale scores),11 the parents in this study scored significantly higher on all three subscales of the MacCAT-CR when compared with the subject populations (table 3). The comparisons of the parents with the published control populations were more variable. On the understanding subscale, three control groups scored higher than the parents (elderly,14 healthy9 and diabetes mellitus11 controls), two scored lower (community members10 and healthy9 controls), and two groups were comparable (elderly and caregivers).13 The parents either scored higher or no different from the published control populations on the appreciation and reasoning subscales, with the exception of the diabetes mellitus controls (table 3).

Subscales were tested for internal consistency and reliability using Cronbach's alpha, with the following results: understanding 0.80; appreciation 0.73 and reasoning 0.92. Choice was not tested as it consists of only one item. Cronbach's alpha computed for all items to check the internal consistency of the MacCAT-CR scale as a whole was 0.86. The Cronbach's alpha was calculated to be 0.90 for the MRI study. Similar indices for the genetics and HRV studies were not estimable due to insufficient variability of scores. The correlation between the understanding subscale and the MacCAT-CR total score was 0.98. A relationship of this magnitude is quite high and suggests that 96% of the variability in the total MacCAT-CR score was explained by the score on the understanding subscale (r2=0.96) (table 4). Subscale–subscale correlations were in the moderate range (0.51–0.67), whereas the subscale–total correlations were in the moderate to high range (0.67–0.98).

Table 4

MacCAT-CR scale–subscale correlations (n=35)


Parents of neonates with critical congenital heart disease demonstrated reasonable understanding, appreciation and reasoning about research participation as measured by the MacCAT-CR. Their subscale scores were significantly better than published values from various adult subject populations, especially those with decreased mental competency or psychiatric dysfunction, and were generally comparable to published adult control groups with some variability (table 3). Overall, parents of neonates with critical congenital heart disease were able to make adequately informed decisions about research participation in spite of the duress of impending cardiac surgery. Our findings appear to be contrary to previous studies, which report that parents of critically ill neonates cannot make valid informed decisions.5 One possible explanation for this difference is the timing of the assessment in relation to the parental decision, with the MacCAT-CR being administered in our study within 10 days of surgery. One survey conducted 18 months later found that 12% parents did not recall making a decision to participate in a randomised controlled trial. Of the remaining 88% who recalled participating in the research, over three-quarters of parents were satisfied with the decisions made.6 Parental satisfaction with the decision to participate in research may be altered by their neonates' health or surgical outcome. Although it is likely that parents may not remember much of the information about a study after 1 or 2 years, the impression of the quality of the decision may be more resilient. It is for these reasons that we believe that the adequacy of parental permission and informed consent should be assessed as close as possible to the decision itself. Our interviews were conducted postoperatively, and may have been affected by immediate surgical outcomes.

Understanding is a critical concern when assessing informed consent. The MacCAT-CR is clearly weighted towards the assessment of informational comprehension, with over 95% of the total score being attributable to the understanding subscale (see table 4). Although the usefulness of the measure in providing a full picture of the non-cognitive aspects of decision-making is thus limited, the parents interviewed for this study appear capable of an adequately informed choice in spite of the emotional duress of the situation. Similarly, Burgess et al20 reported in a retrospective study that 90% of parents felt they made an informed decision, although 38% of parents found recruitment added ‘stress to an already stressful situation.’ We have previously reported that fathers of neonates with critical congenital heart disease diagnosed postnatally had statistically significantly higher levels of both state (current) and trait (baseline) anxiety than fathers of neonates diagnosed prenatally.6 Although one might be concerned about the ability to make decisions under such circumstances, we found no evidence in this small sample of any gender differences in the ability to make an informed decision about research participation in spite of the stress of a critically ill newborn requiring cardiac surgery. Under such circumstances, parents often choose to participate in research studies for societal and potential individual benefit.7

This study is the first report of the use of this validated instrument, the MacCAT-CR, to evaluate parental permission for research participation in parents of critically ill newborns. We found that the majority of parents scored above 76%, chosen to reflect a ‘passing’ score. In particular, only one person each on the HRV and genetics MacCAT-CR scored below 70%. For the more complicated MRI study, five parents scored below 70%. The finding that over 25% of parents did not understand the risks of increasing the duration of anaesthesia and elevating the inhaled carbon dioxide for their critically ill neonate is concerning, despite full disclosure of these risks in the parental permission document. Alternative methods of ascertaining full understanding of the risks before obtaining parental permission should thus be explored. As seen in table 3, the parents in this study had comparable scores to a number of control adult populations in published studies, and better scores than most of the adult subject populations. We found it reassuring that parents who made decisions about research participation in the neonatal period performed as well as, if not significantly better than, members of the general population on understanding and decision-making. There are no data using the MACCAT-CR with other parent populations.

Parental permission for the involvement of critically ill neonates has been shown to be influenced by risk–benefit considerations, compared with other factors such as illness severity or socioeconomic status.21 To facilitate this process, disclosure should be complete, simplified and comprehensible. In addition to receiving the IRB approved permission document, the parents met with an investigator who explained each study in lay language and provided an opportunity to ask questions. One of the limits of our study was that this disclosure process was not evaluated. However, the key content of the permission document was incorporated into the study-specific MacCAT-CR interview tools used in our study, and parents were allowed to answer freely. We were thus able to explore individual components of the permission document, eg, the purpose, procedures, risk, discomforts and benefits of the study. From the understanding that parents displayed, it is probable that the disclosure process for the original study was complete and comprehensible. A recent review of instruments used to evaluate decisional capacity in patients in clinical research settings concluded that the MacCAT-CR had the most empirical support.22 The MacCAT-CR assesses the four domains of understanding information regarding research trial participation, appreciating the relevance of trial participation to the patient's unique situation, reasoning through the presented information to arrive at an expressed choice.

Our study was limited to a relatively small sample size, but was comparable to the samples included in other published studies (see table 3). Our study was not adequately powered to detect significant differences among subgroups of parents. In addition, our conclusions are specific to the three studies, and may not generalise to other research designs such as a placebo-controlled randomised trial. The timing of our interviews after surgery may have impacted the results, because the neonate's outcome was a variable. However, we chose to interview parents as soon after surgery as possible to minimise this impact. Although there was a range of educational backgrounds, with 13 parents having a high school equivalent or less, the sample was relatively homogeneous with respect to ethnicity and race. Our results are limited to a small sample of white, non-Hispanic, well-educated parents. The exclusion of Spanish speakers, which was done to avoid the need for a translated MacCAT-CR instrument and interview, is an additional limitation. The treatment version of the MacCAT has been criticised for an overemphasis on cognitive aspects of decision-making.23 Our study extends this concern to the MacCAT-CR tool, as the major factor in determining the total score is the score on the understanding subscale. This finding is consistent with the recent observation that the level of cognitive functioning was the best predictor of decisional capacity.12 Simply put, the MacCAT-CR does not adequately assess the voluntariness of parental permission or informed consent.

In summary, parents of neonates undergoing cardiac surgery are capable of making an informed choice to permit enrolment of critically ill neonates in research. As a group, their scores are comparable with most published scores using the MacCAT-CR with normal adult control populations. In addition, the MacCAT-CR shows excellent reliability and internal consistency. The MacCAT-CR can be used to assess the informational and decision-making components of parental decision-making about research participation for critically ill neonates. The weighted assessment of understanding suggests that a simpler assessment of parental understanding of information may be able to replace the MacCAT-CR. Further work needs to be done on the non-cognitive aspects of parental decision-making about research participation for critically ill infants and children.


View Abstract


  • Funding Supported in part by a grant from the Florence RC Murray Program of the Joseph Stokes, Jr Research Institute of The Children's Hospital of Philadelphia.

  • Competing interests None.

  • Disclaimer The work reported in this article was performed prior to RMN joining the Food and Drug Administration. The opinions expressed herein do not represent the views of the Food and Drug Administration nor the Department of Health and Human Services.

  • Ethics approval This study was approved by the Children's Hospital of Philadelphia Institutional Review Board.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • The concise argument
    Søren Holm