It is important to review the evidence for and against monitoring infants at home before devising a rational approach to this method of treatment.

At the present time, more neonates and young infants are monitored at home than at any other time in medical history. It is estimated that as many as 15% to 20% of the 400,000 premature infants born annually in the United States are treated using home cardiorespiratory monitoring, primarily for the treatment of apnea and bradycardia, but also for the prevention of sudden infant death syndrome (SIDS). A significant number of children between 1 month and 1 year of age are also monitored because of a history that includes an apparently life-threatening event (ALTE). An ALTE is an unexplained episode in which the observer believes that the life of the child may be at risk. ALTEs are characterized by cyanosis, pallor, cessation of breathing, and loss of consciousness, and are terrifying to the parent or caretaker.

Although many infants are treated using home monitoring, this therapy has not been universally embraced, and it remains controversial. Advocates of home monitoring believe that it can alert caretakers to potentially serious episodes of apnea and/or bradycardia before they cause harm, thus allaying the anxiety of caring for an infant at risk. Opponents of monitoring cite the lack of evidence that these devices substantially alter outcomes for these patients. There is no study, for example, that conclusively demonstrates that monitoring does, in fact, prevent SIDS. Furthermore, many physicians believe that there may be a significant emotional burden associated with monitoring (or deciding not to monitor) a child.

Apnea of Prematurity and Apnea of Infancy
All infants (as well as most children and adults) have apnea, defined as pauses in the breathing pattern. Pathological apnea, therefore, refers to an interruption in breathing of 20 seconds or more (or of less than 20 seconds, when accompanied by a significant change in heart rate or oxygen saturation). Significant bradycardia, for the young infant, consists of a decrease to a level that is at least one third less than the resting heart rate. Since oxygen saturation rarely decreases below 85%, a level lower than that is often considered abnormal. Infants with chronic lung disease of prematurity (bronchopulmonary dysplasia) may need to show a decrease to 80% saturation before it is considered significant, however, because they commonly have lower resting saturations (even when they receive supplemental oxygen). The fall in saturation, in all cases, should last for at least 5 seconds to eliminate the possibility of motion artifact. These definitions, while not universally agreed upon, represent significant apnea, bradycardia, and oxygen-saturation changes that are rarely seen in normal, healthy infants beyond 36 weeks of postconceptional age.

Apnea is commonly classified as central, obstructive, or mixed. In central apnea, all respiratory effort ceases. During obstructive apnea, the infant continues to exhibit respiratory effort, but gas flow to the lungs is prevented by altered structure or tone within the airway. Mixed apnea contains elements of both central and obstructive apnea, whether seen within the same apneic pause or at different times during a period of respiratory recording. All forms of apnea are visually difficult to detect, although obstructive apnea may be more obvious to the trained observer, since the infant may struggle to overcome the airway obstruction. For precise diagnostic purposes, multichannel recordings of breathing (most commonly nasal airflow and thoracic impedance, averaged heart rate, and oxygen saturation) are necessary. It has been shown2 that trained observers will miss more than 50% of episodes of apnea of prematurity (AOP) detected by multichannel recording.

Another form of apnea, periodic breathing, consists of pauses in respiration of less than 10 seconds separated by periods of breathing lasting less than 20 seconds, occurring at least three times in succession. Periodic breathing occurs as the principal form of apnea in approximately 20% of premature infants.2 Whether this type of apnea has the same significance as other forms of apnea is unknown. In the healthy premature infant, periodic breathing rarely occurs during more than 5% of the total sleep time of the infant.3 Periodic breathing should be considered significant if it exceeds that percentage, if it persists for more than 15 minutes, or if it is accompanied by changes in oxygen saturation. Clinical detection of periodic breathing without recording several hours of breathing during sleep is virtually impossible. Controversy about periodic breathing exists because it is found during sleep in healthy individuals. In these cases, however, periodic breathing rarely persists for longer than 5 minutes, and it is not accompanied by saturation changes. There are currently no prospective studies that definitively relate periodic breathing to SIDS or to other adverse outcomes of any type.

Apnea occurs in approximately 70% of infants at 34 to 35 weeks of postconceptional age.1 While AOP decreases progressively in frequency over the subsequent weeks of life, approximately 30% to 40% of infants born prematurely will still demonstrate AOP at their expected due dates (40 weeks after conception). Significant apnea and/or hypoxemic events have been well documented beyond this age in premature infants.4-6 The mean age of resolution for AOP is approximately 50 weeks after conception.7 Other recent studies6,8 indicate that 6% to 22% of infants of very low birth weight (VLBW) have apnea at term, and that 91% of premature infants have apneas of more than 12 seconds’ duration at the time of hospital discharge. Of these infants, 31% also have bradycardia and 6.5% require prolonged hospitalization because of the severity of their apnea and bradycardia.6 It is therefore evident that AOP does not resolve at term in many VLBW infants and may be present for some time after hospital discharge.

In term infants, apnea is seen less frequently. Data from our nursery indicate that about 2% to 3% of term infants have demonstrable apnea when studied in the immediate postpartum period. The long-term significance of this apnea is unknown. There are no data to prove that term infants with apnea immediately after birth are at greater risk of ALTE or SIDS than other term infants. Only 10% of infants who present with an ALTE during the first months of life lack an antecedent history of apnea, indicating the highly unpredictable nature of all of these events during the first months of life.

Apnea, SIDS, and Home Monitoring
A recent Pediatrics editorial stated that “the SIDS field of research is still bogged down by apnea monitors, which have proven of little value. Hundreds of millions of dollars have been wasted over the last 25 years. Worse still, the device continues to be used. The apnea monitoring business has become a religion. More people are living off SIDS than dying from it.”9

This statement reflects the great frustration that many physicians have felt with deaths from SIDS occurring among their patients and the apparent inability of home monitoring to prevent these deaths. In reviewing the available data, however, it is apparent that the outcome is not necessarily so negative.

For term infants, it has been acknowledged for some time that the peak age of death due to SIDS occurs at 2 to 4 months of age, at a mean of approximately 52 weeks postconceptional age. The peak age of death for the premature infant is approximately 4 to 6 months, but at a similar mean of 52 weeks after conception.10 The premature infant, therefore, may have a longer risk period for SIDS following hospital discharge. There are few data that compare risk in preterm infants born at different postconceptional ages with respect to SIDS. The similarity in postconceptional age of SIDS death for both term and preterm infants, however, suggests the possibility that some neurodevelopmental phenomenon is one of the etiologies of this as-yet-unexplained problem.

Although premature infants comprise approximately 10% of the birth population, they account for slightly more than 20% of SIDS deaths.9 Because immaturity of respiratory control is so commonly seen in the prematurely born infant, it has been suggested that there is a relationship between AOP and the risk of SIDS. In neonatal intensive care units (NICUs), premature infants frequently stop breathing unexpectedly, often with bradycardia and oxygen desaturation. In many instances, it appears that the infant would not resume breathing without direct intervention. Consequently, the hypothesis of apnea as an initiating event for SIDS is attractive, since the premature infant does not struggle to resume breathing (making the situation similar to that seen in many cases of SIDS). To date, however, this theory of causation for SIDS has not been established. Furthermore, most full-term children who die of SIDS (accounting for 80% of SIDS deaths) do not appear to have apnea prior to death, from parental reports. As noted, however, visual detection of apnea and periodic breathing in most instances is very difficult, even for medical personnel, and parents may miss such episodes.

Most unfortunately, there does not appear to be any simple, accurate method for predicting which infants are likely to die of SIDS.11 Many programs have been developed, however, to provide home cardiorespiratory monitoring for infants in an effort to reduce the incidence of SIDS.12 To date, no prospective, randomized, controlled study has proven that home cardiorespiratory monitoring will prevent SIDS. Anecdotal data from some programs suggest that the incidence of SIDS can be reduced in a population of monitored infants, but it has not been shown that it is home monitoring itself that causes this reduction in SIDS incidence.11

In our apnea program, we have not encountered the death of a premature infant in nearly 7 years, during which time we have discharged more than 3,000 infants. Based on currently available statistics, we would have expected at least 5 to 6 deaths during that time. Among our term infants, only one child has died during the past 7 years (and under unusual circumstances, since the infant had evidence of massive hemolysis just prior to death). To obtain this very low death rate, however, we monitored approximately 45% of the premature infant population for an average of 4.5 weeks, as well as 20% of term infants referred to our program for ALTE. Opponents of monitoring would argue that similar results could have been obtained without monitoring. To achieve these results, however, it might be necessary to use inpatient hospital observation for days or weeks. The cost of hospitalization has risen to such an extent that a single inpatient day may be as expensive as 3 to 4 months of home cardiorespiratory monitoring. It would, therefore, appear that the judicious use of home monitoring in this era of cost consciousness is a reasonable approach, even if only some deaths can be prevented. Why, then, has home cardiorespiratory monitoring achieved such a negative reputation?

Home Monitoring Pros and Cons
In order to evaluate the criticism of home cardiorespiratory monitoring, it is necessary to determine the goals of this therapy. Furthermore, it is also essential to examine similar uses of cardiorespiratory monitoring elsewhere in medicine in order to evaluate its effectiveness.

Cardiorespiratory monitoring has long been the cornerstone of intensive care (especially in NICUs). Virtually all infants receive surveillance with a variety of monitors in these units, yet the effectiveness of this costly activity has not been well substantiated. Most neonatologists have seen children die while on hospital monitors, or require vigorous resuscitation because a monitor failed to alert the caregiver to an adverse event. More than 50% of apnea/bradycardia events in the hospital environment are missed, yet NICU monitoring would never be dismissed as a waste of resources. Furthermore, failure to provide such care would most likely constitute a breach of the standard of care, opening hospitals and physicians to substantial litigation. How can the same therapy exist as a standard of care in one set of circumstances, yet be viewed as an inappropriate use of resources in other circumstances, when identical patient goals (the detection and prevention of adverse cardiorespiratory events) are present? The answer is that it cannot be viewed in this way; the problem with home monitoring is that expectations for it are unrealistic.

Physicians and nurses who practice in NICUs understand that monitoring will occasionally have its failings; they act accordingly, providing enhanced bedside vigilance for their most fragile patients. In NICUs that offer extracorporeal membrane oxygenation (ECMO), for example, it is not uncommon to see two nurses, or a nurse and ECMO technician, at the bedside because of the known risks of this procedure. If monitoring were failure proof, such intense bedside care would not be needed. In the home situation, however, any failure of monitoring becomes unacceptable, since it may result in death. The loss of a single infant who has been monitored for SIDS typically leads to a condemnation of the practice of monitoring in its totality. Furthermore, because SIDS is so rare an event, a single death may dramatically alter statistics, making it extremely difficult to demonstrate the value of monitoring. Total condemnation of monitoring, however, seems unwarranted, and is potentially unfair to parents and infants.

When a child is born prematurely and has apnea and bradycardia, or when a child presents with an ALTE, the physician is faced with a very difficult, complicated decision: should the child remain hospitalized and, if so, for how long? Medical staff often refer to “killer apnea episodes.” This is unquestionably a poorly chosen term, but it is one that accurately depicts many events. Many children stop breathing, dramatically drop their heart rates, and respond only to active intervention (and, at times, vigorous resuscitation). The child never appears to struggle and the situation is otherwise peaceful, except for the physician, nurse, or parent who witnesses the event. These individuals are panic-stricken at the prospect that the child might do the same thing at home. Some recent data would suggest that if a child has no further episodes for 8 days or more, it is safe to discharge him or her.13 Other studies, however, dispute these data and indicate that apnea—especially in premature infants—may persist for some time after the due date, placing the child at risk for a much longer period.3-8 The choice that the physician then faces is whether to continue hospitalization or to discharge the infant with a monitor.

When apnea is present, there is usually a point at which the episodes are relatively rare and easily stopped, with the intervention usually consisting of simple tactile stimulation of the infant. Consequently, many physicians elect to use home monitoring rather than prolong hospitalization. They recognize the limitations of the technology, particularly for children with obstructive apnea, but have made a reasonable medical decision that is based on a simple concept: the episodes of apnea are so infrequent and so mild that care can probably be given at home, as long as the parents have some type of support. For this support, monitoring is generally used; otherwise, many families will actually take shifts at the bedside to watch their infant. In the era prior to the

use of home monitoring, this practice was seen with some degree of regularity. Furthermore, as noted previously, monitoring is far less costly than continued hospitalization. With 2 additional days in the hospital being as expensive as 6 months of home monitoring, it simply does not make sense, in most cases, to keep the child in the hospital. While some unscrupulous physicians may attempt to profit from the use of monitoring, most are simply making a rational decision that would appear to be in the best interests of the child and the family and to keep costs to a minimum.

Some parents decide not to use the monitor in spite of medical advice, others may fail to hear an alarm, and still others are not sufficiently skilled to resuscitate their child. Children do, therefore, die while being monitored, no matter how well home care is planned. The rare SIDS death that does occur while the infant is monitored then leads to a complete condemnation of the practice of home monitoring. The problem, however, is that the extent of the death rate without monitoring is unknown. The opportunity to perform a prospective, randomized study of home monitoring is long gone because of the passions evident on both sides of the monitoring issue.

Prior to the 1992 institution by the American Academy of Pediatrics of supine sleep for young infants, which dramatically lowered the SIDS rate in the United States, the rate was already declining. The reasons for this decline were not clear, but the trend was surprising, given the increased survival of VLBW infants during the preceding decade, since they are statistically more prone to SIDS. The use of exogenous surfactant, improved ventilatory techniques, and a variety of other factors might all be responsible, but the increasingly widespread use of home monitoring would seem to be a possible factor as well. Furthermore, even if SIDS is not a consideration, the role of monitoring in neonatal outcomes may be an issue. Recent data14 have indicated that children with grades 3 and 4 intraventricular hemorrhage may have more significant apnea and oxygen desaturation than other premature infants. It is not known whether the reduction of apnea, through monitoring and the use of pharmacological agents such as methylxanthines, improves neurodevelopmental outcomes in premature infants. Because of the complex nature of the medical problems that are experienced by preterm infants, it is difficult to estimate the contribution of apnea to any developmental disability. Frequent apnea, especially when accompanied by bradycardia and oxygen desaturation, may adversely affect neurodevelopmental outcomes and should be reduced or eliminated, if possible. Additional work is required to establish this relationship and to determine whether monitoring and the reduction of apnea can improve neurological outcomes.

Monitoring may also be an effective diagnostic tool for use in the homes of children who are having events that are so sporadic that they cannot easily be detected in the hospital.15,16 With documented monitoring, these infrequent episodes of apnea or ALTE can be detected and quantified, while allowing a more normal home situation for the child. If events are detected, additional evaluation for a specific cause may be initiated. Studies such as electroencephalography, electrocardiography, central nervous system imaging, and metabolic studies may be valuable in such circumstances.

Many neonatologists are aware of the substantial anxiety that exists at the time of discharge for any family taking home an infant of low birth weight.17 With recent changes in inpatient care, many premature infants now leave the hospital at a weight of 1.8 to 2 kg.18 The discharge of a tiny infant may be very stressful for the family, and home monitoring may be a helpful adjunct in the transition of the infant to the home setting. Any caretaker who has a child on home monitoring must be carefully trained, both in the use of the monitor and in cardiopulmonary resuscitation. Parents should always be clearly informed, however, that monitoring might not prevent SIDS.

The goals of home monitoring are:

  • to minimize hospitalization and the cost of care for infants with apnea and/or ALTE;
  • to provide reasonable parental reassurance and security upon hospital discharge, especially for VLBW infants who are discharged at lower weights and younger postconceptional ages;
  • to permit the optimal neurodevelopmental outcome, unaffected by apnea and oxygen desaturation; and
  • to serve as a diagnostic tool for home events.

There are several types of monitors available for home use. The most common form of monitoring combines impedance pneumography with averaged heart rate to provide cardiorespiratory monitoring. Recently introduced units also are capable of event recording and storage as well as compliance detection, for the evaluation of both home events and the frequency of monitor use. Many units now have modems for the transmission of data to the physician for evaluation. Other devices, such as pulse oximeters, piezo belts, and pressure capsules, have either been impractical to use or exhibited limitations in their application. Newer technologies and software programs, however, may make such devices as oximeters more practical in the near future.

All devices are associated with alarms that occur when no true cardiorespiratory event is present. Such false alarms are often worrisome to parents, who may discontinue monitor use if they occur too frequently. Usually, excessive false alarms can be minimized through alternative placement of electrodes combined with parental education and support. In addition, event recordings from home monitors allow the evaluation of these problems. This valuable addition to care can eliminate many needless days of hospitalization. Few physicians would disagree with the fact that children do not belong in hospitals unless it is absolutely necessary.

With any use of monitoring, careful follow-up care is indicated. If the physician caring for an infant has limited experience with home monitoring or cannot interpret monitor recordings, assistance from a center or program that has such expertise is indicated. The mean duration of home monitoring for most infants should usually be less than 6 to 10 weeks. Extended monitoring should be reserved for those infants who continue to demonstrate significant cardiorespiratory abnormalities. Only in the rarest of circumstances should any child be monitored beyond 1 year of age. Most often, children who require monitoring in such circumstances have other technological dependencies (such as bronchopulmonary dysplasia requiring home mechanical ventilation). Monitoring to prevent SIDS beyond 1 year of age is not indicated.

Increasing evidence from many countries indicates the importance of two primary factors for SIDS prevention in term infants: placing children supine for sleep and avoiding exposure of the fetus and infant to cigarette smoke.19 Unless there are clear contraindications, infants should be placed to sleep in this manner, with smoking avoided both during pregnancy and in the infant’s environment after birth. These interventions appear to be most significant in their ability to reduce SIDS. Monitoring, however, may be valuable in some clinical situations, both for the treatment of apnea and to assist families in coping with the care of their infants. 

Alan R. Spitzer, MD, is a professor of pediatrics at Thomas Jefferson University, Philadelphia. Beginning in October 1999, he will be professor of pediatrics and chief of the division of neonatology at State University of New York, Stony Brook.


1. Southall DP, Levitt GA, Richards JM, et al. Undetected episodes of prolonged apnea and severe bradycardia in preterm infants. Pediatrics. 1983;72:541-551.

2. Spitzer AR, Fox WW. Infant apnea. Pediatr Clin North Am. 1986;33:561-582.

3. Stefano JL, Anday EK, Davis JM, Fox WW, Spitzer AR. Pneumocardiograms in healthy premature infants—a study of normative longitudinal data. Am J Perinatol. 1991;8:170-173.

4. Poets CF, Stebbens VA, Alexander JR, Arrowsmith WA, Salfield SA, Southall DP. Arterial oxygen saturation in preterm infants at discharge from the hospital and six weeks later. J Pediatr. 1992;120:447-454.

5. Samuels MP, Poets CF, Southall DP. Abnormal hypoxemia after life-threatening events in infants born before term. J Pediatr. 1994;125:441-446.

6. Barrington KJ, Finer N, Li D. Predischarge respiratory recordings in very low birth weight newborn infants. J Pediatr. 1996:129:934-940.

7. Lipsky CL, Gibson E, Cullen JJ, Cullen JA, Spitzer AR. When does apnea of prematurity resolve? Pediatr Res. 1993;33:267A.

8. Eichenwald EC, Aina A, Stark AR. Apnea frequently persists beyond term gestation in infants delivered at 24 to 28 weeks. Pediatrics. 1997;100:354-359.

9. Lucey J. Editorial. Pediatrics. 1999;103:812.

10. Lipsky CL, Gibson E, Cullen JA, Rankin K, Spitzer AR. The timing of SIDS deaths in premature infants in an urban population. Clin Pediatr. 1995;34:410-414.

11. Shannon DC. Prospective identification of the risk of SIDS. Clin Perinatol. 1992;19:861-869.

12. Spitzer AR, Gibson E. Home monitoring. Clin Perinatol. 1992;19:907-928.

13. Darnall RA, Kattwinkel J, Nattie C, Robinson M. Margin of safety for discharge after apnea in preterm infants. Pediatrics. 1997;100:795-801.

14. Cheung PY, Barrington KJ, Finer NN, Robertson CM. Early childhood neurodevelopment in very low birth weight infants with predischarge apnea. Pediatr Pulmonol. 1999;27:14-20.

15. Weese-Mayer DE, Silvestri JM. Documented monitoring: an alarming turn of events. Clin Perinatol. 1992;19:891-906.

16. Silvestri JM, Hufford DR, Durham J, et al. Assessment of compliance with home cardiorespiratory monitoring in infants at risk of sudden infant death syndrome. J Pediatr. 1995;127:384-388.

17. Singer LT, Salvator A, Guo S, et al. Maternal psychological distress and parenting stress after the birth of a

very low-birth-weight infant. JAMA. 1999;281:799-805.

18. Spinner SS, Girifalco RB, Gibson E, Stavis RL, Greenspan JS, Spitzer AR. Earlier discharge of infants from neonatal intensive care units: a pilot program of specialized case management and home care. Clin Pediatr. 1998;37:353-357.

19. Henderson-Smart DJ, Ponsonby AL, Murphy E. Reducing the risk of sudden infant death syndrome: a review of the scientific literature. J Paediatr Child Health. 1998;34:213-219.