Some of the first follow-up programs for premature infants were questionnaire surveys undertaken in Europe in the second decade of the twentieth century. Not surprisingly, these surveys did little to clarify the long-term prognosis of premature infants. By the 1950s and 1960s, some significant methodologic advances had occurred: population-based samples and standardized assessments. With the advent of newborn intensive care units in the early 1960s, the population of low-birth-weight survivors began to grow, and by the late 1970s, newborn follow-up programs had begun to proliferate.

The need for follow-up programs for high-risk newborns has now been widely accepted. However, there has been little discussion of their multiple (and sometimes competing) purposes, which have included providing comprehensive and specialized care for high-risk survivors, monitoring rates of handicap among infants discharged from intensive care units, and answering specific research questions.

In designing a follow-up program, a careful delineation of its specific purposes is key. This delineation will inform decisions about issues such as patient selection, frequency of visits, evaluation methods used, interventions used, communication with parents and community physicians, duration of follow-up, and program funding.

Two contrasting approaches to follow-up care are outlined in Table 28.1. The clinical service model is designed to provide individualized medical care to high-risk survivors. This approach has the advantages of flexibility, efficiency, and orientation to patient needs. Although widely used, this approach is unlikely to provide unbiased estimates of the prevalence of health and development problems in surviving high-risk infants, or an assessment of the effects of perinatal problems on and therapies for various outcomes. Such information has considerable relevance for health policy and perinatal care. The epidemiologic model is designed to address these additional needs. However, the epidemiologic model entails considerable effort and expense to identify and recruit all high-risk patients in a defined region, to train staff, and to ensure unbiased, valid, and well-calibrated assessments of outcome.

For most follow-up programs, some synthesis of the clinical service and epidemiologic approaches is needed to ensure that both follow-up care and outcome assessments are of high quality. Some outcome measures (e.g., developmental assessments at specific age points) should be assessed by personnel carefully trained and blinded to previous findings. To reduce the likelihood that examiner expectations for high-risk infants would influence the evaluations (and to provide a full-term frame of reference), the examiners should know that some control infants (e.g., healthy infants born at term) will be included. At the same time, health care should be given on a schedule determined by patient needs, and it should be provided by personnel who are familiar with perinatal and follow-up findings and who have a good rapport with the family. Whenever feasible, follow-up should be provided for infants with any of the neonatal problems listed in Box 28.1.

For infants who have primary pediatricians skilled in managing disorders such as bronchopulmonary dysplasia, short-gut syndrome, and the sequelae of intracranial hemorrhage and hypoxic-ischemic encephalopathy, follow-up clinics may need to provide no more than standardized neurodevelopmental assessments. However, indigent populations have the most limited access to health care and the highest prevalence of neurodevelopmental morbidity. For other infants, the provision of comprehensive medical care in the follow-up clinic may be important, both to manage health problems and to reduce loss to follow-up. If more than 10% to 20% of patients are lost to follow-up, only imprecise and possibly biased estimates of the prevalence of handicap will be obtained. The fact that traditional follow-up programs are not designed to provide comprehensive health care partly accounts for the uncertainty about the outcome of high-risk infants from economically disadvantaged families. Broyles et al. conducted one program of comprehensive care for high-risk inner-city infants. In that program,
the addition of comprehensive primary care (instead of conventional follow-up care only) yielded reductions in three areas: loss to follow-up; total health care costs; and, most important, the incidence of life-threatening illnesses (death or admission for pediatric intensive care).