Missed and delayed diagnoses of CCHD
For decades, standard clinical practice has called for examining the cardiovascular system as part of the routine newborn examination. Despite this practice, however, one in three infants with a potentially life-threatening form of CHD is discharged from the hospital nursery undiagnosed.4
Clinical evaluation with prenatal ultrasound or newborn physical examination alone can easily miss CCHD, particularly in infants with subtle clinical signs.5 For instance, half of all newborns with CCHD—particularly those with ductal-dependent defects—have no distinctive murmur.6 In many cases, symptoms of CCHD do not present until after hospital discharge, contributing to high rates of morbidity and mortality for newborns with CCHD.7
Missed diagnoses prevent newborns with potentially life-threatening cardiovascular malformations from receiving successful treatments now available with advances in pediatric cardiology and cardiac surgery. Furthermore, although options for surgical correction may be the same regardless of whether the diagnosis is made early or late, the effects on treatment outcomes are substantial.6
Delaying treatment until newborns with CCHD are critically ill is associated with increased surgical mortality, prolonged hospital stay, and higher incidence of such serious adverse effects as neurologic dysfunction, including learning difficulties and delays in cognitive development.6 The goal of screening is to detect CCHD prior to clinical deterioration in affected newborns.
Despite efforts to increase the recognition of CCHD, timely diagnosis remains an elusive goal. Missed and delayed diagnoses of CCHD occur in anywhere from one in 3,500 to one in 25,000 live births, depending on the type of defect.4,8-10 Chang and colleagues evaluated trends in missed diagnoses of CCHD in a population-based study of California statewide death registry data collected between 1989 and 2004.11 Although the detection of CCHD improved between 1989 and 1999, the total annual number of infants with missed or late diagnoses of CCHD was unchanged between 2000 and 2004. Overall, up to 30 infants died each year in California due to missed or late diagnoses of CCHD. The median age at death was 13.5 days, underscoring the importance of early diagnosis.11
The consequences of missed diagnoses have also been observed in real-world practice. In a retrospective analysis of newborns younger than age 30 days admitted to the Children’s Hospital of Philadelphia with CCHD, 6.7% had a significant physiologic compromise, including metabolic acidosis, seizure, cardiac arrest, or renal or hepatic injury, due to a missed diagnosis.12
The persistence of missed CCHD diagnoses highlights system-based deficiencies in diagnostic protocols for newborns. For instance, major changes to cardiopulmonary physiology occur during the first few days of life, making diagnosis difficult during this period. Moreover, the trend toward earlier hospital discharge—and the associated shift of newborn care to other health-care settings—adds to the challenge of timely diagnosis. Indeed, the setting where CCHD is first recognized correlates with clinical outcomes, with earlier diagnosis a strong predictor of lower morbidity and mortality.13
Detection of hypoxemia: visual observation vs. pulse oximetry
Visual observation. Determining whether an infant is pink or blue involves examining the face, trunk, and mucous membranes to detect the presence of central cyanosis. However, several factors other than oxygen saturation can influence the infant’s color. Patient factors such as skin thickness and color, perfusion, and hemoglobin concentration can influence color, while environmental factors such as ambient light conditions can influence color perception.14 Therefore, a major limitation of the newborn physical examination is the inability of the human eye to detect important degrees of cyanosis.15 The gap between normal oxygen saturation and visible cyanosis has been described as the “cyanotic blind spot” (Figure 1).16
The ability to detect mild cyanosis by visual examination and the reliability of visual perceptions across different observers are poor even among neonatal intensive-care personnel.
One recent study showed substantial variations in the clinical assessment of infant color.14 In the study, 27 physicians and nurses were asked to observe video recordings of 20 infants taken immediately after delivery. The clinicians were asked to indicate whether each infant was pink at the start of the video, became pink during the recording, or had never been pink. Only one infant (5%) was perceived to be pink by all clinicians. For the remaining 19 infants, clinicians varied widely in their opinions of newborn color. Between 4% and 81% believed that the remaining infants never became pink. Even when clinicians agreed that infants turned pink, the SpO2 (arterial oxyhemoglobin saturation as measured noninvasively by pulse oximetry) at which individual infants were perceived to become pink varied from 10% to 100%. Furthermore, the false-positive for hypoxemia was high, with 17% of infants rated as “never pink” despite maximum SpO2 levels ≥95%.