Objective. To define current practice among US newborn screening programs for notification of results, research, and consenting procedures.
Methods. A telephone survey of all US newborn screening program supervisors.
Results. All 51 programs participated. All states reported abnormal results to the infant’s physician, and some also reported to the hospital and parents. Cases with abnormal results were tracked to different endpoints but usually (92.1%) at least until a follow-up appointment was made. A total of 66.6% of programs can communicate with programs in other states; 9.8% enable families to suppress reporting of results to the infant’s physician. No state has a mechanism for parents to prevent results from entering the medical record. Parents or physicians who request results are often authenticated by providing their name (52.9%). Many programs (45.1%) report only to physicians and require just their name (43.5%), an identification number (17.4%), a letter (26.1%), or a parent’s signature (26.1%). A total of 70.6% retain residual blood samples; of these, only 8.3% store them completely devoid of patient identifiers. A total of 49.0% of programs aggregate data for research. In 16.0% of these, the data are publicly available. In 24.0%, researchers obtain approval at their own institution; in 24.0%, researchers obtain approval through the state laboratory Institutional Review Board. In 74.5% of programs, parents are notified but not asked for consent before collection of the sample; 19.6% neither notify parents nor obtain consent before screening.
Conclusions. There is wide variation in practice among the US newborn screening programs. Because the programs collectively manage a comprehensive nationwide genomic databank, careful consideration of how information technology and high-throughput genomic analysis are used will be essential to allow progress in clinical care, public health, and research while protecting individual privacy.
State and regional programs screen the approximately 4 million infants who are born in the United States each year for a cluster of genetic conditions.1 Testing is done on blood obtained by heel stick and stored on pieces of filter paper. Results of many of the tests performed on the filter paper samples must be communicated to patients and physicians emergently so that additional testing, dietary modification, or other treatments may be initiated. Because these filter paper blood samples are stored beyond the neonatal period, newborn screening programs collectively manage a genomic databank for the younger segment of the US population.2 Information about an infant’s genetic variation may be of substantial interest to parents and physicians but also to researchers, health insurers, and marketers.
We conducted a survey of the nation’s newborn screening programs to define the current range of practice for consenting, notification, and research. An understanding of current procedure will help form the basis for development of the policies and public health infrastructure proposed in the Newborn Screening Taskforce.3
A survey of screening laboratories was conducted by telephone in July and August 1999. Respondents were supervisors of newborn screening programs or state laboratories in each of the 50 states and the District of Columbia. They were identified through program Web sites or through telephone calls to the state departments of public health. The supervisors were sent a letter describing the project and then called by 1 of the authors (S.F.). The Children’s Hospital Committee on Clinical Investigation approved the research protocol. Verbal consent was obtained from each respondent before the interview began.
The survey consisted of 34 questions and took 10 to 15 minutes to administer. Topic areas included consenting procedures, notification, case tracking, confidentiality, authentication, storage of blood samples, and research policies. Survey responses were entered into a relational database, and results are reported descriptively as frequencies. The data are reported in aggregate without identifiers for individual states.
Supervisors from all of the state laboratories and the District of Columbia responded to our telephone survey, making our total sample size 51.
Policies for notification of the screening test results differ in all states depending on whether the result is normal or abnormal. Two states do not notify at all if results are normal. The remaining states notify the birth hospital (44 of 49), the infant’s doctor (27 of 49), the infant’s parents (1 of 49), or some combination thereof. The most frequent means of reporting normal results is by letter (48 of 49); fax, e-mail, and computer disk are each used by 1 state. The time frame for notification of a normal result ranges from immediately after the result is obtained up to 1 month later (Fig 1).
The variation in procedures for notification of abnormal results is much wider. The combinations of methods used—telephone call, letter, certified letter, and fax—are shown in Table 1. The recipient of the abnormal results varies as well. All states report to the infant’s physician with some also reporting to the birth hospital, the infant’s parents, and/or a geneticist (Table 2). Twenty-six states use a uniform method for reporting results, and 25 states vary the method according to criticality of the results. For borderline results, 21 of 25 states send a letter to the infant’s doctor, 7 of 25 send a letter to the birth hospital, and 7 of 25 send letters to both the birth hospital and the infant’s doctor. When results are strongly positive, states most often contact the doctor—24 of 25 via telephone, 10 of 25 via letter, 9 of 25 via both letter and telephone call, 6 of 25 via fax to the doctor, and 6 of 25 via telephone call and fax.
In some states, all of the newborn screening responsibilities, from analyzing the samples to notifying the appropriate people of the results, lie with the state laboratory. However, in other states, a newborn screening follow-up program receives all results from the state or contract laboratory, communicating and tracking them. Thirteen states provided information about how results are transferred between the laboratory and the newborn screening program. The most common method used is telephone call (8 of 13) followed by e-mail and telephone call (2 of 13), e-mail only (1 of 13), telephone call and fax (1 of 13), and letter (1 of 13).
Nearly all of the states (47 of 51) have a procedure in place to follow up the reporting of an abnormal test result: 1 state tracks until an appointment has been made, 1 tracks until an appointment has been kept, 23 states track until diagnosis has been either confirmed or denied, 22 states track until treatment has been initiated, and 9 continue periodic follow-up after treatment has been initiated. Five states terminate tracking once the results have been given to the infant’s doctor.
The newborn screening programs are able to track cases through contact with the infants’ physicians. Nineteen states receive notification from the doctor who treated the infant—12 by letter and 7 by telephone call. Nineteen other states make a follow-up telephone call to the doctor. Thirty states track the time elapsed from when results first become available to when families, doctors, or hospitals are first notified, and 38 track the time from when results first become available to when the infant is treated.
Some families are difficult to locate. Fifty of 51 states do not make additional efforts to track down the family if the results are normal, although 1 state will mail a letter if the family relocated to a different state and left a forwarding address. When the results are abnormal, 34 of 51 newborn screening programs will contact the newborn screening program in another state if the family gave a forwarding address. Otherwise, the newborn screening program staff will speak to family contacts (14 of 51); send a public health nurse to the family’s last known address (10 of 51); mail a certified letter to the last known address (6 of 51); or contact the department of health, registry of motor vehicles, and use any possible means to find the infant (11 of 51). Three states report never having had trouble locating a family.
Of great concern for newborn screening and other genetic testing is confidentiality of test results: who will see them, how they will be stored, and who has access to them. With regard to the primary use of the newborn screening data—the detection of treatable newborn disorders—there is often very little recourse for parents who want to keep results private. In 46 states, the birth hospital will receive a hard copy of the test results, and in 33 states the infant’s doctor will receive one. Very few (5 of 51) states have a mechanism whereby parents can suppress the reporting of results to their infant’s physician. In 4 states, although there is no formal mechanism, consideration will be given on a case-by-case basis to parents’ requests for suppression of newborn screening results, and in 1 state the parents must sign a form requesting that their infant’s physician not be notified. In no state is there a formalized mechanism whereby a parent can completely prevent the reporting of newborn screening results in the infant’s medical record.
Authentication of the Patient
With regard to patients’ obtaining their own results, there is wide variation in the policies of different states concerning how test results are requested, what modalities may be used to request results, what information must be provided about the patient, and what identifying information the person requesting results must provide. In 1 state, test results may not be requested. In the other states (Table 3), test results may be requested by telephone (43 of 51), letter (40 of 51), fax (34 of 51), e-mail (18 of 51), or voice response system (6 of 51). All states require the infant’s name to give results, but they all also require at least 1 additional piece of information (Table 4), such as the infant’s date of birth (47 of 50), birth hospital (25 of 50), infant’s mother’s name (29 of 50), infant’s mother’s address (7 of 50), infant’s mother’s Social Security number (9 of 50), infant’s mother’s telephone number (4 of 50), infant’s birth weight (1 of 50), medical record number (4 of 50), or infant’s laboratory identification number (9 of 50).
Authentication of Other Requesters
States often have several methods for authenticating the people who request results (Table 5). Twenty-three newborn screening programs allow test results to be given only to physicians who request the information, whereas 27 states allow information to be given directly to parents who request it. However, in 7 of those states, the parent must have the result interpreted by the infant’s doctor or hospital, whereas in 20 states, someone at the screening program interprets the results. Five states use a voice response system in which each doctor must use his or her unique personal identification number. In 29 states, a parent or a doctor who requests results must state his or her name, whereas in 10 states, not only must a parent or doctor state his or her name, but also the newborn screening program will then call back to the telephone number listed for said person to verify the identity. Twelve states will release results when they receive a request written on official letterhead, and 6 states will report results to a physician with a signed release from the infant’s parents; 2 states will release results to the parents with a signed release. Four states will give results to a physician who is known to the newborn screening program. One state requires a copy of a Social Security card or a driver’s license. Twenty-four of the 50 states that allow results to be requested keep track of who has requested screening results, and 25 track to whom they have been reported.
Research and Public Health
Storage of Blood Samples
Thirty-six of the 51 newborn screening programs retain the residual blood samples, at least for a short time. Of these, 31 store samples with the individual patient identifiers still on them, 2 of 36 store them completely deidentified, and 3 of 36 use a coding system that prevents the researcher from knowing the identity of the patient but that can be decoded should the need arise. Although 12 states store blood samples for no officially stated reason, the remainder of the states store blood samples for specified purposes, such as future testing (13 of 36), testing requested by the family in a case in which the child dies (7 of 36), quality control to check errors in testing (8 of 36), or research (5 of 36). In 25 states, aggregate data from the newborn screening program are used for research.
The approval processes for a researcher to be given access to these aggregate data differ. In 4 of 25 states, researchers need not seek specific permission to use these data, as aggregate data are open to the public. Researchers must go through the Institutional Review Board at their own institution in 6 of 25 states and through an Institutional Review Board at the state laboratory or newborn screening program in 16 of 25 states. Researchers must submit a request detailing the proposed research to the director of the newborn screening program in 5 of 25 states, and the senior staff members of the newborn screening program review each specific research proposal in 2 of 25 states.
The approval processes for public health officials to gain access to these data are the same: aggregate data are publicly accessible (6 of 25), health officials in a department other than the newborn screening program must go through their own Institutional Review Boards (3 of 25), health officials must go through an institutional review board at the state laboratory or newborn screening program (14 of 25), health officials must submit a request to the director of the newborn screening program (4 of 25), and the senior staff members at the newborn screening program review the request (2 of 25).
In most programs (38 of 51), parents are notified but not asked to give consent for collection of the sample for mandatory testing. Ten states neither notify parents nor ask for consent in screening their newborns. Before collecting specimens from newborns, only 3 states require parents’ signed consent. In 2 states, parental refusal of newborn screening is not permitted, but in the majority of states (27 of 51) parents may, for religious reasons, refuse screening. Parents may verbally refuse screening for religious reasons in 5 states and for any reason in 1 state. In 15 states, parents must sign a waiver refusing screening.
High-throughput testing for genetic disease is becoming widely disseminated, as is the information technology that will make the testing results accessible. Duly authorized parties as well as others will likely have access. Screening programs will need to evolve to integrate these technologies. We found a wide variation in practice patterns among the newborn screening programs in the 50 states and the District of Columbia, findings that update and complement previous work.4
Screening programs seem not yet to be fully prepared to 1) optimally leverage information technology to report results to patients and physicians and 2) manage a population-based genomic databank while addressing issues of human subjects and patient confidentiality.
The Newborn Screening Taskforce recognizes that the “environmental context within which these programs were established has changed dramatically over the past 10 years”3 and has issued an extensive set of recommendations. These recommendations include developing new procedures for consenting patients, inventing new testing protocols, establishing a public health infrastructure that connects newborn screening programs to the medical home, and increasing the role of the family in the entire spectrum of the process.
Networked databases5 with Web-based front ends would improve the notification practices by enabling increased access to newborn screening results as well as other personal medical data. The increased access to information by caregivers and families could aid considerably in notification, case tracking, and follow-up. The CORN report, “US Newborn Screening System Guidelines II: Follow-up of Children, Diagnosis, Management, and Evaluation,”6 recommends that every normal result be sent to the infant’s health care provider within 7 days and that failure of receipt of that result trigger an action. Networked databases have the capability also to improve communication among programs when infant residences change. Our results show that this sort of communication does not occur consistently, with only two thirds of states having formal mechanisms to contact other state screening programs. Not until databases and automation are in place will such result be achievable.
Authentication procedures for physicians and patients who request data have been appropriately weak to maximize access to results of critical medical importance; most programs give out results to physicians who call with a limited set of patient demographic information. As testing programs expand to include results of a less critical nature, the patient should be in greater control of the release of personal genomic data.7 Approaches to informed consent necessarily must evolve8,9 as newborn screening programs broaden from public health service providers to custodians of a population-based genomic databank that is used for research.10 Currently, most states do not allow families to suppress the reporting of results to physicians, nor do they require consent for testing or storage of samples. Biological samples unlinked to identifiers have traditionally been made available for research without Institutional Review Board approval. There are 2 important considerations when dealing with these types of data. First, anonymization of medical record databases—stripping them of patient identifiers so that clinical data can be aggregated for research—has limitations. Data sets stripped of most identifiers often have a high vulnerability to reidentification through merges with other publicly available databases.11–13 Second, genomic data form a fingerprint of sorts, and people can be identified from the data themselves. These problems with reidentification must be accounted for when developing procedures for consent and anonymization.
We found evidence of wide variation in practice patterns among the newborn screening programs in the 50 states and the District of Columbia. Although additional work is needed to relate the variability to issues of quality and ethical conduct, these data help to inform the process of implementing the Newborn Screening Task Force recommendations. Because the newborn screening programs collectively manage a nationwide comprehensive genomic databank, careful consideration of how information technology and high-throughput genomic analysis are used is essential to allow progress in clinical care, public health, and research while protecting individual privacy.
This work was supported, in part, by the National Library of Medicine Next Generation Internet Initiative Contract N01-LM-9-3536.
- Received April 10, 2001.
- Accepted July 11, 2001.
- Reprint requests to (K.D.M.) Division of Emergency Medicine, Children’s Hospital, 300 Longwood Ave, Boston, MA 02115
- ↵Council of Regional Networks for Genetic Services (CORN) Newborn Screening Committee. National Newborn Screening Report—1995. Atlanta, GA: Council of Regional Networks for Genetic Services;1999
- ↵Newborn Screening Task Force. Serving the family from birth to the medical home: a report from the Newborn Screening Task Force convened in Washington DC, May 10–11, 1999. Pediatrics.2000;106(suppl) :383– 427
- ↵Pass KA, Lane PA, Fernhoff PM, et al. US newborn screening system guidelines II: follow-up of children, diagnosis, management, and evaluation. Statement of the Council of Regional Networks for Genetic Services (CORN). J Pediatr.2000;137(4 suppl) :1– 46
- ↵Mandl KD, Szolovits P, Kohane IS. Public standards and patients’ control: how to keep electronic medical records accessible but private. BMJ.2001;322 :283– 287
- ↵Sweeney L. Replacing personally-identifying information in medical records, the Scrub system. Proc AMIA Annu Fall Symp.1996;333– 337
- Copyright © 2002 by the American Academy of Pediatrics