It is not clear why differences in end-of-life decision making for infants and children in similar circumstances, except for their place of residence, exist. As with other intensive care practices, the goal should be to identify and implement best practices. However, objective data to guide these decisions, such as predictive models of who will survive mechanical ventilation, or survive without severe and permanent injury, remain poorly developed. Indeed observational studies suggest that ICU physicians discriminate between survivors and non-survivors more accurately than do scoring systems in the first 24 h of ICU admission, implying limited usefulness of outcome prediction in the first 24 h for clinical decision making (Sinuff et al. 2006). Moreover, end-of-life care differs from most other medical problems, in that best practice not only is defined in terms of evidence-based strategies but also relates to prevailing cultural, religious, and legal norms (Cook et al. 2006; Truog 2005).

A decision-making process about the withholding or withdrawal of mechanical ventilation for infants and children at the end of life that meets the approval of all cultures across the world is unlikely. Yet, most would likely see a decision-making process about end-of-life care that always leads to only one conclusion as harmful. For example, a process that always advocates for the withholding or withdrawal of life-sustaining treatment for critically ill infants or children would rightfully be seen as harmful and flawed for failing to recognize instances where providing ongoing treatment was in the child’s best interests. Similarly, a process that never allowed for the withholding or withdrawal of life-sustaining treatment for critically ill children would also be seen as harmful and flawed for failing to recognize instances where biologic existence supported by medical technology may be technically possible but may not be in the infant or child’s best interests. The healthcare personnel involved in non-initiation and withdrawal of care may also be concerned about lack of or limitations of legal provisions to protect them. The laws related to withdrawal of care are not uniform across nations and often nonexistent in many countries (Rao 2005). The decisions about withdrawal or non-initiation should consider the best available and recent data for the outcomes of the disease with the available type of medical care in the given circumstances. Understandably as the quality of medical care improves, the outcomes may change. For example, traditionally, babies who develop severe IVH have been believed to have high mortality and high incidence of sequelae in survivors. Recent studies have, however, shown a decrease in mortality from severe IVH to 30–40 % and markedly improved quality of life in survivors (Bassan and du Plessis 2008; Bassan et al. 2007; Roze et al. 2008). Hence, a decision to actively withdraw support from a hemodynamically stable premature infant with severe IVH may be difficult to justify ethically in centers with good quality medical care. Withdrawing ventilator assistance also involves consideration of stopping other treatments. Caregivers and family may feel greater dilemmas with regard to withholding feeds and fluids than the withdrawal of mechanical ventilation. In the developed world, withholding medical fluids and food may be morally more acceptable (Porta and Frader 2007). The American Academy of Neurology sanctions the withdrawal of nutrition and fluids in persistent vegetative states if agreed by their care providers and parents (Position of the American Academy of Neurology on certain aspects of care and management of persistent vegetative state patient 1989). In the Eastern cultures, however, the families are more likely to be inclined to continue food and water even by artificial means. Giving food and water to the hungry and thirsty is considered a great act of compassion, while starving a dependent person is abhorred (Singh 2007). However, all will agree that aggressive nutritional interventions like parenteral nutrition can be withdrawn (Carter and Leuthner 2003) (Fig. 73.1).

To address the need for a framework to guide the decision-making process, a growing number of national professional societies have issued guidelines around decisions to forego life-sustaining treatment for critically ill infants and children (American Academy of Pediatrics Committee on Bioethics 1994; Aita et al. 2008; Bulow et al. 2008; Cabre et al. 2008; Cremer et al. 2007; Hubert et al. 2005; Mani 2006; Pedersen et al. 2007; Tibballs 2006; Truog et al. 2005; Waisel et al. 2002; Zamboni 2002).