Performance audit. Department of Human Resources, Genetic Newborn Screening Program / Performance Audit Operations Division, Department of Audits and Accounts

PERFORMANCE AUDIT
Department of Human Resources
Genetic Newborn Screening Program
November 2002
Russell W. Hinton, State Auditor Performance Audit Operations Division 254 Washington St, SW Department of Audits and Accounts Atlanta, GA 30334

The Genetic Newborn Screening Program was created in response to a 1978 statute that

mandated the Department of Human Resources (DHR) create a system to test all Georgia

newborns for specific genetic disorders that can be lethal in the early days of a newborn's

life. The lethal affects of each of these disorders can be

prevented if they are rapidly detected and treated. These relatively rare genetic conditions include metabolic, endocrine, and sickle cell blood disorders. Metabolic and endocrine disorders cause imbalances of body chemistry, which can lead to mental retardation and death if left untreated. Sickle cell blood disorders make newborns highly susceptible to fatal infections,

Immediate medical management of an infant with one of these disorders will prevent many, and in some disorders all, of the serious clinical [consequences].

and require ongoing antibiotic treatment and pain management. The Journal of Pediatrics

A full description of each of these disorders is included in

October 2000

Appendix A on page 33.

According to state statute, the genetic services system for the detection and treatment of these disorders is to include: screening newborns for the disorders; following-up on those who screen positively to ensure
diagnostic testing is conducted (note that the screening process casts a wide net within which only a small number of infants will actually have a disorder); obtaining a specific diagnosis; initiating and continuing appropriate treatment; and assessing the program.

Genetic Newborn Screening Program

Page 1

Timeline of when Disorders were added to Genetic Newborn Screening in Georgia

1965 1970 1975 1980 1985

1966 Phenylketonuria
(PKU)
1972 Sickle Cell Blood
Disorders (targeted population screening)
1978 Galactosemia Tyrosinemia Maple Syrup Urine
Disease (MSUD) Congenital
Hypothyroidism Homocystinuria

1990

1989 Congenital Adrenal
Hyperplasia (CAH)

1995

2000 2002

1998 Sickle Cell Blood
Disorders
(all newborns)

Source: Official Code of Georgia Annotated

NOTE: See Appendix A (p. 33) for a detailed description of each disorder, its impacts, symptoms, and treatments.

In 1966, Georgia passed legislation requiring all newborns be tested for phenylketonuria (PKU), a metabolic disorder; however, no centralized program was established to monitor the screening and follow-up process. With the 1978 initiation of a comprehensive genetic newborn screening system, five additional disorders were added for statewide screening. Since then, the law has been amended to now require that all newborns be tested for eight specified disorders (see the timeline on the left). The statute also provides for DHR to add similar disorders that cause mental retardation, as it deems necessary.
The Program is administered within the Office of Infant and Child Health Services in the Family Health Branch of the Division of Public Health in the Department of Human Resources (DHR). As shown in the chart on the right, program activities are conducted by the Georgia Public Health Laboratory (GPHL); contracted follow-up service providers: Emory University Department of Pediatrics, Division of Medical Genetics, Grady Memorial Hospital Comprehensive Sickle Cell Center, Medical College of Georgia Sickle Cell Center, and the Sickle Cell Foundation of Georgia; and Program Management within DHR. Additionally, hospital personnel and physicians are involved with sample collection and treatment services. (This process is illustrated in the flow chart on page 4.)
The Program's management of the screening process is facilitated by a monthly meeting of the Program's "Work Group," which consists of Program staff, representatives from each of the contracted entities, and representatives from the Georgia Public Health Laboratory (GPHL). These meetings are used to discuss Program procedures, problems encountered, and future efforts. The Work Group makes recommendations to the Newborn Screening Advisory Committee, which was created in response to a 1999 federal program

Genetic Newborn Screening Program

Page 2

review that advised the Program to create a committee of experts to discuss policy initiatives for the Program. The Advisory Committee meets quarterly, and consists of Program representatives (including follow-up contractors' staff, GPHL personnel, and the Program Manager), legal advisors, representatives from advocacy groups, medical professionals, a hospital representative, and parents/consumers. This group does not have official written by-laws to govern their composition and attendance or to delegate specific authority. However, the Advisory Committee does make recommendations to the Council on Maternal and Infant Health (M&I Council), which functions in an advisory capacity for DHR's Board of Human Resources. (See the following diagram for an illustration of this process.) State law officially recognizes the M&I Council, and its members are appointed by the Governor.

Exhibit 1

Structure of the Genetic Newborn Screening Program

Board of Human Resources

Department of Human Resources (DHR)

Council on Maternal and Infant Health

Division of Public Health (DPH)

Family Health Branch

Georgia Public Health Laboratory (GPHL)

Office of Infant and Child Health

GNBS Advisory Committee
Key

Genetic Newborn Screening Program
(GNBS)

Organizational relationship
Advisory relationship

Contracted follow-up care providers: Medical College of Georgia (MCG) Grady Memorial Hospital (Grady) Emory University Pediatric Genetics
(Emory) Sickle Cell Foundation of Georgia
(SCFG)

Source: Program literature, contracts, and state statute

As illustrated and described on the following pages, the screening process includes: collecting blood samples; testing all blood samples for the disorders; and follow-up notification, diagnosis, and treatment. According to state statute, this process, through the point of diagnosis, is to be completed by the time the infants are 21 days old.

Genetic Newborn Screening Program

Page 3

Exhibit 2
Screening

Blood Sample Collection Hospital or
physician's office collects a newborn's
blood sample

Screening
Blood Sample Collection DHR Regulations require hospitals to collect blood samples from all infants born under their

Blood sample is mailed to and received by the
Georgia Public Health Laboratory (GPHL)

Unsatisfactory samples

care. The samples are collected by sticking babies in the heel and absorbing the blood into special laboratory testing paper on the sample collection form. Regulations specify that these samples be collected 48 hours after birth or at the time of an

Sample satisfactory for
testing?

Letter requesting new No sample sent to hospital
and/or physician

infant's discharge from the hospital, whichever is earlier. If an infant is discharged before 48 hours

Yes
Sample Testing GPHL tests the blood sample for evidence of 8 different genetic
disorders

These samples are currently tested, as well. Their results are only reported if they are positive. Another sample is necessary for truly accurate testing.

of age, regulations require a second specimen be collected no later than when the infant is one week old, and places responsibility on the hospital administrator, or his/her designee, to give the parents written and verbal instructions to have the second sample taken. A second specimen is

Results Reporting Positive metabolic and endocrine test results are: sent to Emory (GPHL calls Emory
regarding panic-level results)

required in these cases because certain proteins must build up in the baby's blood to a detectable level before screening can be accurately

All test results are: mailed to the hospital & physician, entered in the Voice Response
System, and sent to the Program Office

completed. When infants are born outside of a hospital or birthing center (at home, for instance), regulations require the person in charge of the

facility or the person in attendance, to give

written notification...of the legal requirements for

Positive screening?

No No further action

the newborn to be tested. By law, parents can only refuse screening based on religious grounds.

Follow-up

If parents chose to do so, a hospital official is to

Yes

inform the parent of the consequences of refusal...

metabolic or endocrine disorder?
No
sickle cell disorder?
No
sickle cell trait carrier?

Results Transmittal Yes Emory receives positive results from the GPHL

Results

Transmittal

Yes

Program Manager sends positive

results to MCG and

Grady for follow-up

and require the parent to complete a statement

Notification Follow-up
contractors attempt to notify physicians
and/or parents regarding each infant's positive
screening

Diagnosis Follow-up contractors attempt to obtain a confirm atory diagnosis on each infant positively screened

Confirmation in Care
Follow-up contractors ensure each infant affected with a disorder is in
care

indicating their declination of newborn screening

Program Manager sends results to Sickle Cell Foundation of Georgia
Yes SCFG sends letters notifying mothers of the trait carrier
screening & offers counseling

for religious reasons. Records of refusal are to be retained by the facility.

Genetic Newborn Screening Program

Page 4

Blood Sample Testing and Release of Results Blood samples are sent directly to the Georgia Public Health Laboratory (GPHL) for testing. The GPHL conducts screening tests for all eight disorders and is responsible for distributing the test results. Samples that are damaged or collected poorly can be inadequate for obtaining accurate test results, and are designated as "unsatisfactory samples" by laboratory staff. Current laboratory procedures require these samples be tested anyway; however, a second sample is requested so accurate testing can be completed. Test results on the unsatisfactory sample are only released if they are positive. GPHL retains the blood samples for six weeks and retains the test results for two full calendar years following testing. Test results are mailed to the physician listed on the specimen collection form (i.e., the "physician of record") and the hospital that collected the sample.

All doctors can choose to register with the GPHL to receive access to the Voice Response System (VRS), an automated telephone system for obtaining test results. The Program's VRS was designed to ease physician's access to test results. It allows physicians to phone in, enter a passcode, and obtain infants' screening results as well as determine if blood samples have been received by the laboratory. Physicians enter their assigned passcode and the mother's Social Security Number (SSN) or the unique "lab form number" located on the blood sample collection card to obtain results. (The lab form number currently is only accessible to the doctor if the specimen was collected at his/her office.) It should be noted that the laboratory is in the process of issuing new blood sample forms that will provide mothers with a copy of the lab form number.

Follow-up

If a sample screens positively for a disorder, the infant must undergo a follow-up diagnostic test to confirm the disorder and begin any necessary treatments. The Program has contracted with the Emory University Department of Pediatrics, Division of Medical Genetics (Emory); Grady Memorial Hospital's Comprehensive Sickle Cell Center (Grady); the Medical College of Georgia Sickle Cell Center (MCG); and the Sickle Cell Foundation of Georgia (SCFG) for follow-up services. Emory is responsible for follow-up on all infants positively screened for any of the seven metabolic and endocrine

According to The Journal of Pediatrics, the primary function of the follow-up component is to locate infants with positive screening results and to facilitate the entry of these infants into the diagnostic and management components of the NBS system in a timely fashion.

disorders. MCG and Grady geographically divide responsibility for

follow-up on infants positively screened for sickle cell disorders; and SCFG provides

follow-up on all sickle cell trait carriers. Follow-up includes notifying parents and

physicians regarding positive screening results, obtaining diagnostic tests on all positively

screened newborns, and providing necessary counseling and care.

Emory newborn screening staff daily access positive screening results for any of the seven metabolic and endocrine disorders through an electronic link to the GPHL's database. If a

Genetic Newborn Screening Program

Page 5

screening result suggests "panic" level values for a disorder, meaning immediate medical care is necessary, GPHL personnel call Emory staff with the results to speed the notification process. Positive results for sickle cell disorders are sent electronically from the laboratory to the Program Office on a daily basis. The Program Office in turn sends these results to Grady and MCG. Until May of 2002, the results were sent to the two entities by facsimile on a daily basis. Recently, the Program implemented an electronic transfer of information to Grady and MCG on a twice-weekly basis. Under the new system, the contractors are able to annotate infants' records, allowing Program management to identify what follow-up actions have been taken. The Program reports that it plans to implement a similar system for Emory, which would allow for review of Emory's follow-up activities, as well.

Metabolic and Endocrine Disorders--Follow-Up Notification, Diagnosis, and Treatment Emory contacts the hospital, physician of record, or the parent (if the physician cannot be located), to arrange for confirmatory testing at Emory's diagnostic center. In certain instances, infants' initial screening results and general health status can indicate that a repeated screening test conducted by the GPHL is sufficient to confirm they do not have a disorder and, therefore, no diagnostic test is necessary. This determination is made by Emory medical staff. Infants diagnosed with a disorder can receive medical treatment through the Emory Pediatric Medicine Genetics Clinic. Emory staff also give their recommendations for treatment to private physicians who wish to test or treat an infant independently.

Sickle Cell Disorders--Follow-Up Notification, Diagnosis, and Treatment Grady handles follow-up notification for sickle cell affected infants residing in the Atlanta area. MCG handles notification for infants born in all other parts of the state. Each entity is responsible for contacting the hospital, physician of record, or the parent (if the physician cannot be located) to ensure the infant undergoes a confirmatory diagnostic test and begins necessary medical care. Both Grady and MCG also utilize public health departments to locate infants. Additionally, MCG's laboratory runs confirmatory diagnostic tests for any child in the state at no cost to the physician or parent. However, doctors may choose to have the tests run elsewhere. In contrast to metabolic and endocrine screening tests, positive sickle cell blood disorder screening results always indicate the presence of a blood abnormality or that the infant is a trait carrier. However, specific typing of the disorder is necessary for appropriate treatment and counseling.

Information on infants identified as carriers of sickle cell trait is electronically transferred from the Program Office to the Sickle Cell Foundation of Georgia (SCFG) on a weekly basis. The SCFG mails infants' screening results to mothers along with information regarding the condition, instructions on how to obtain confirmatory testing if desired, and the option to receive free counseling from the SCFG. The laboratory at MCG will also run confirmatory tests for sickle cell trait at no cost to the physician or parent. There is no specific treatment necessary for sickle cell trait carriers; however, people with sickle cell

Genetic Newborn Screening Program

Page 6

trait should be aware of possible problems under certain low oxygen conditions, such as during deep sea diving or in surgery, and of the chances of having children with the trait or a sickle cell disorder. SCFG tracks the total numbers of trait carriers contacted and reports them to the Program on a semi-annual basis.

In 1999, a program review was conducted by a nationally recognized team of newborn screening experts sponsored by the U.S. Department of Health and Human Services. The review was conducted at the request of the Division of Public Health; and a document of findings was presented to the Division at the conclusion of the review. This review was the first and only of its kind. The Program has instituted some of the recommendations, including developing monthly Work Group meetings and creating the Advisory Committee. Further efforts recommended in the review, and supported by this audit, are included in the Recommendations and Agency Responses section of this report.

The GPHL reported receiving 202,577 blood specimens in calendar year 2001. However, no data was available regarding the number of individual babies screened since multiple specimens were submitted for some infants. By comparing screening and birth records from 2001, the audit team estimated that at least 123,936 (91.4%) of the 135,617 infants born in Georgia in 2001 were screened. The number of infants actually screened could be higher since significant data entry errors and/or omissions in either the birth or screening records could have resulted in records not matching, and infants not born in Georgia also could have been screened. The audit team's comparison of birth and screening records is explained more fully in the Recommendations.

Program staff reported that in calendar year 2001, 5,632 infants were referred to Emory for medical follow-up. Of these, Emory reported successfully locating 5,597, and reported 97 diagnosed cases of clinically significant metabolic and endocrine disorders. Due to missing or incorrect identification information, 35 newborns could not be located (i.e., were lost to follow-up). In these cases, parents may have purposely or accidentally provided incorrect information, may have changed names, addresses, or phone numbers following the birth, or hospital and/or GPHL staff may have incorrectly entered the contact information into the laboratory's database.

The Program reported that 143 infants were referred to MCG and Grady for sickle cell disorder follow-up in calendar year 2001. The Sickle Cell Foundation of Georgia (SCFG) reported that 9,682 infants were referred for having tested positively as sickle cell trait carriers. SCFG contacted 9,627 of the affected infants' parents by letter, reporting that 55 were lost to follow-up.

In calendar year 2000, the Program coordinated educational seminars to train hospital personnel in proper blood specimen collection and handling procedures. Seminars were

Genetic Newborn Screening Program

Page 7

held in four locations throughout the state. The Program has initiated another round of seminars for 2002. The Program also has an instruction manual for specimen collection, an instructional video for hospital and medical staff, and a brochure for parents on the genetic screening program.

Georgia does not charge a fee for newborn screening or for the follow-up notification,

diagnostic testing, and medical care provided as part of the screening system. Through state

general funds, the Federal Medicaid Matching Grant, and the Federal Maternal and Child

Health Block Grant, DHR manages the Program, conducts

sample testing, and contracts for follow-up care. Additional Program-related costs are incurred by hospitals and doctors' offices collecting the samples, by follow-up entities, and by health departments that aid in follow-up.

Cost estimate for Genetic Newborn
Screening Activities in 2001
Total cost: $3.1 million Cost per infant screened: $25 Cost per positive diagnosis: $12,894

The cost of the Program for fiscal year 2001 is estimated at $3.1 million. The pie charts (Exhibit 3, below) illustrate the distribution of these funds among the Program's activities, and indicate the fund sources for the $3.1 million. This figure translates into a cost of approximately $25 per screening test run, and of approximately $12,894 per confirmed sickle cell, metabolic, and endocrine diagnosis in calendar year 2001 (of which there were 239 total).

Exhibit 3

Estimated Genetic Newborn Screening Costs Breakdown for Fiscal Year 2001

Breakdown of Costs by Newborn Screening Activity
Total costs $3.1 million

Breakdown of Costs by Funding Source
Total costs $3.1 million

Laboratory screening tests
47417/.25%%

Program administration

Federal Medicaid
Matching Grant
27 297/1.09%%

4.3% 4 3/10 %

114.8/5%%

Federal Block Grant

771./25%%

Follow-up notification and diagnostic tests
48418/.52%%

Additional Sources
(i.e., costs covered by

contracted entites

through other means)

State Funds
62624./85%%

Source: The estimates provided were derived from costs provided by the GPHL, Program staff, and follow-up contractors. No estimate of the total cost of operations was available from the Program.

Genetic Newborn Screening Program

Page 8

The overall estimate does not include the efforts and costs incurred at the local level by public health departments, or by the hospitals and private practitioners in collecting the specimens. Additionally, the cost assessment only includes follow-up through the point of contact with a medical professional caring for the affected child and diagnostic testing completed by the follow-up contractors. Treatment costs, such as the costs for metabolic formula, are not included. The cost for initiation of treatment for infants affected by the metabolic, endocrine, and blood disorders varies according to the child, the disorder, and the affect of the disorder on the child. For instance, infants affected by the disorders may need immediate hospitalization upon diagnosis, or need to be started on special formula, medicine, or a diet upon a visit to a doctor's office. Hospital fees and physician costs likewise vary. It should be noted, however, that the state does provide limited funding for long-term care for metabolic, endocrine, and sickle cell disorders. Additionally, insurance, third party payers, and Medicaid may cover the costs of treatment.

Exhibit 4

21 12

9 10 9 10 14 9 27

More than 8 Disorders 8 Disorders

7 Disorders 6 Disorders 5 Disorders 4 Disorders

U.S. Newborn Screening
Mandated Disorders July 1, 2002

3 Disorders

Source: National Newborn Screening and Genetics Resource Center

All states have genetic newborn screening programs, but they vary in the disorders they screen for, what oversight responsibilities exist, the processes for informing parents, allowed exemptions, specimen storage policies, and use of blood samples and payment for newborn screening procedures. However, it is estimated that these programs collectively result in the screening of approximately 4 million infants in the United States each year.

There are approximately 700 genetic tests available; but not all of the tests have been

Genetic Newborn Screening Program

Page 9

proven to be beneficial to infants. As a result, each state has selected those disorders for which it will screen. The disorders chosen for screening in each state can vary for multiple reasons, including differences in population that can lead to different incidence rates of disorders in the populations, making them more or less desirable for a public health screening program. Since all genetic screening programs are state mandated and operated, the amount and type of follow-up care provided also varies by state.

Every state routinely screens for PKU and congenital hypothyroidism. In contrast, Georgia is only one of four states currently screening for tyrosinemia. In some states, additional tests may be conducted by the state at the parents' discretion. The map in Exhibit 4 on the previous page illustrates the number of disorders for which each state currently screens.

In addition, all but eight states currently charge parents a fee for the mandatory screening. Georgia is among those in which screening is conducted free of charge. The testing fees in other states range from $10.50 to $59.50. It should be noted, however, that the number of newborns, the technology in use, and the number of tests being conducted can all affect the cost for screening. For instance, some states have started using a tandem mass spectrometer (MS/MS) to test for many disorders. This machine can be used to test for multiple disorders simultaneously, but it is costly to purchase and maintain, and it requires a great deal of expertise to operate.

Because of the complexities involved in testing, follow-up, and treatment, and the differing political and economic environments across the states, operation of newborn screening varies widely from state to state.
Brad Therrell, Ph.D., Director of the National Newborn Screening and Genetics Resource Center
(In a hearing before the U.S. Senate Subcommittee on Children and Families, June 2002)

The primary objective of this audit was to determine if DHR is fulfilling its responsibility to create a system for the detection and treatment of genetic metabolic, endocrine, and sickle cell blood disorders in all Georgia newborns. The audit was conducted in accordance with generally accepted government auditing standards for performance audits and focused on calendar year 2001. The audit methodology included reviews of Program and follow-up contractor records, interviews with Program staff in the Division of Public Health and the follow-up contractors' offices, a data match of birth and newborn genetic screening records from calendar year 2001, observations of Work Group and Advisory Committee meetings, and a review of Best Practices through published literature and interviews with nationally recognized experts in the field of genetics and genetic newborn screening. This report has been discussed with appropriate personnel representing the Department. A draft copy was also provided for their review and comment. Pertinent responses received from the Department have been incorporated into this report as appropriate. The audit scope did not include: review of specific laboratory techniques; analysis of medical advice or care provided by the follow-up contractors; or detailed review of the follow-up provided to sickle cell trait carriers by the Sickle Cell Foundation of Georgia.

Genetic Newborn Screening Program

Page 10

Recommendation No. 1
The Department should establish sufficient managerial review and internal controls to ensure the Program is fulfilling its purposes. Currently, the Program cannot provide reasonable assurance that: all Georgia newborns receive a screening test; followup care (including a diagnostic test and treatment) is provided to all newborns who require it; or programmatic operations are assessed on an ongoing basis. As discussed below, improvements are necessary in all phases of the Program's operations. These points are discussed in more detail in subsequent recommendations.
The Program cannot ensure that all newborns receive a valid screening test. No assurances exist that all children born in Georgia receive a screening, or that sufficient action is taken to ensure all samples are satisfactory for testing.
The Program cannot provide assurances that all newborns who screen positively for a disorder receive follow-up care, or that the care is provided in a timely manner. In addition, no internal controls have been established to ensure that test results are transmitted accurately, completely, and in a timely fashion to the follow-up care providers so that these children can be located.
There is currently no ongoing assessment of the Program as required by state statute. Program staff do not have or collect information on the costs to conduct the Program, the time frames in which screening is

conducted, or an evaluation of the disorders for which the state should be screening.
The Division should make improvements in all aspects of the Program, and its oversight, to ensure the Department is in compliance with its statutory mandate to screen all newborns and ensure they receive necessary follow-up contact, diagnosis, and treatment within three weeks of life. These points, and others, are discussed in more detail in the following recommendations. Additionally, specific legislative matters and other informational findings are included at the end of the report.
The Department of Human Resources (DHR) indicated in its response to this report that it plans to undertake several initiatives to improve the overall operations of the Genetic Newborn Screening Program. Among the initiatives noted were: recruitment of new personnel to monitor hospitals' screening of newborns on a quarterly basis; continued and expanded training of hospital staff to improve the satisfactory rate of submitted blood specimens; and centralization of follow-up data in a new system to be developed in approximately 18 months. Examples of the Department's current efforts to address concerns raised in this report are included in the following recommendations. It should be noted, however, that additional corrective actions are necessary to fully address the recommendations of this audit.

Genetic Newborn Screening Program

Page 11

Recommendation No. 2

The Program should implement an

ongoing assessment to identify those

newborns not screened and take

corrective action, thereby ensuring all

newborns are screened as required by

law.

There is currently no attempt to determine on an ongoing basis if all newborns are screened. Nor are there efforts to ensure that those newborns not screened are identified and contacted for timely screening.

Georgia statute makes the Department responsible for the screening of all newborns. However, there is currently no

attempt to

determine on an ongoing basis if all are

screened. Nor are there efforts to ensure

that those newborns not screened are

identified and contacted for timely

screening. The severity and rapid impact of

these disorders requires timely screening

and a completed diagnostic test within three

weeks of life to prevent the lethal effects.

The audit team conducted a comparison of calendar year 2001 birth and screening records obtained in February 2002 to determine if unscreened infants could be identified in a timely manner and contacted for screening. The comparison identified 11,681 individual newborns (8.6% of the 135,617 babies born in 2001) who could not be linked to a screening record.1 A DHR epidemiologist estimated that even a comparison every five days using this or a

similar methodology would allow for the identification and contact of a large number of the unscreened newborns in a timely manner.

In 2001, a federal grant application required that the Program demonstrate the proportion of the newborn population screened. The Program conducted a comparison of 1999 birth and screening records, which indicated a 97% screening rate in that year. However, this review was not completed until 2001, and no attempt was made to retrieve the remaining 3% for screening. In addition, Program staff stated that no review has been conducted for any other year; and the Program cannot provide assurances of the proportion of newborns that are being screened.

It also should be noted that the Program

could take steps to evaluate individual

hospital's screening practices as a

mechanism to ensure all newborns are

screened. Of the 11,681 infants born in

2001 who could not be linked to a

screening record, 19% (2,189) were born in

three metro Atlanta hospitals. Were the

Program to

institute a review of the hospitals in which unscreened infants were born, it could target its oversight efforts and potentially

The Program has no record indicating the number of infants screened. Nor could the Program provide us with any assurances of the proportion of newborns that are being screened.

(1) It should be noted that there are several reasons a newborn may not be screened. For instance, a newborn may rightfully not be screened: if he or she dies in the hospital before 48 hours of age; if the parents have a religious exception to the screening; or if the infant was born outside of the state, but the birth was registered in Georgia. The number of infants not screened for these reasons could not be quantified. In addition, significant data entry errors and/or omissions of data in either the birth or screening records could have resulted in records not matching.

Genetic Newborn Screening Program

Page 12

remedy problems that result in newborns not being screened.
Such comparisons of birth and screening records--if conducted in a more frequent and timely fashion, and followed by screening of the identified infants-- could both provide the Program with a safety net to ensure all newborns receive a screening, and provide a mechanism to evaluate the reasons newborns are not currently screened. Additionally, hospitals could be targeted for training or additional oversight by the Program.
The DHR stated in its response that it will conduct matches of birth and screening records every three months to identify unscreened infants. A new position will be created to identify the unscreened newborns and conduct follow-up to ensure they are subsequently screened. (Note that state statute requires that the screening process through the point of diagnosis be completed by three weeks of age.) DHR will target hospitals with larger numbers of unmatched birth and screening records for immediate visits by public health staff.
Recommendation No. 3
Action should be taken to ensure an adequate sample is received for all newborns so that a complete and accurate screening test can be completed. When the laboratory deems a newborns' blood specimen unsatisfactory for adequate testing, the Program does not take sufficient action to ensure a satisfactory replacement sample is received. (The

reasons a specimen may be deemed unsatisfactory are listed in the box below.) Laboratory procedures require that a letter be mailed to the hospital

[A] mechanism should be in place to ensure that a satisfactory specimen is received within a reasonable time. Additional action should be taken until a satisfactory specimen is received or the case is designated lost to followup.
The Journal of Pediatrics October 2000

and the

physician of record notifying them of the

unsatisfactory nature of the original sample

and stating that a replacement sample is

necessary for accurate testing. The

Program, however, does not take any

action to ensure a second sample is

submitted. The audit team reviewed 59

unsatisfactory samples. Even after the

laboratory's notification, no satisfactory

replacement sample was submitted to the

laboratory for 29 (49%) of these infants.

Reasons the Laboratory may deem a Blood Specimen Unsatisfactory
(Note: The blood specimen is collected by soaking the blood into a filter paper form.)

Quantity of blood was not sufficient to test (including no blood on the form)
Too much blood over saturated or the blood was unevenly soaked into the filter paper
More than 7 days lapsed between collection and laboratory receipt
Alcohol, water, or another substance contaminated the blood
Filter paper was roughed up, as by a blood collection instrument
Crumpled form Expired filter paper form Filter paper detached from demo-
graphic data form

Genetic Newborn Screening Program

Page 13

The Program conducted no further followup to ensure that another specimen was collected. For the 30 unsatisfactory specimens that were replaced, it took from 6 to 230 days (with a median time of 32 days) from the infants' birth for the laboratory to receive the replacements. By 32 days of age infants are already at risk for the lethal impacts of these disorders. In addition, state statute requires the process from birth through diagnosis be completed in 21 days. Given the early onset and detrimental effects of these disorders, more aggressive action should be taken to ensure an accurate test can be conducted.
Best practices also recommend that unsatisfactory samples be treated with the same urgency as positive screening results. By the Program's standards of follow-up, this would mean phone calls to the physician, hospital, and county health department, until the infant could be located. The 1999 federal review of the Program also emphasized the importance of instituting a more active system of unsatisfactory [specimen] follow-up. The review pointed out that 7.7% (14,597) of the specimens submitted in 1997 were unsatisfactory. If not replaced with satisfactory specimens, these represented possible missed diagnoses for 3-4 Congenital Hypothyroidism, 1 PKU, and 79 sickle cell disease effected infants.1 In calendar year 2001, the laboratory reported receiving 17,510 unsatisfactory samples, representing 8.64% of the 202,577 samples received. It should be noted that Program

staff agree that mailing a letter to the doctor and hospital for a replacement sample is ineffective because the baby typically is no longer at the hospital once

7.7% of the specimens submitted in 1997 were unsatisfactory. If not replaced with satisfactory specimens, these represented possible missed diagnoses for 3-4 hypothyroid, 1 PKU, and 7-9 sickle cell disease effected infants.1
1999 federal review of the Program

the letter arrives, and the physician listed

may not be seeing the baby (e.g., because

the doctor listed was the physician on call

at the time of birth or the mothers'

obstetrician).

In order to decrease the total number of unsatisfactory samples received by the laboratory, the Program has plans to regularly notify hospitals of their unsatisfactory sample rate, the reasons their samples were unsatisfactory, and how to improve specimen collection and transport. Such notification was in the planning stages for the duration of this audit; and the first reports were sent to hospitals in August 2002. The Program also should establish a procedure to aggressively seek replacement samples and verify that they have been received.

The Program stated in its response that in addition to mailing the hospital reports noted above, it also plans to compare each hospital's specimen quality to the state and ideal rates. Based on these reports, technical assistance will be provided to the

(1) In its response, the Department noted that the laboratory's practice in 1997 was to test unsatisfactory specimens; and the federal review's projection does not account for disease-affected infants that might have been identified through this testing. See Recommendation No. 9 regarding the continued testing of unsatisfactory specimens.

Genetic Newborn Screening Program

Page 14

hospitals. In addition, the Program is planning ways in which to notify hospitals in a more timely manner that an unsatisfactory specimen has been received. The Department stated that future efforts will include monitoring to ensure these samples are replaced with satisfactory specimens.
Recommendation No. 4
The Program should institute a comprehensive process for determining whether all positively screened newborns receive appropriate follow-up care. Currently, the Program cannot provide assurances that all newborns who screen positively receive follow-up care. Although the Program receives monthly and annual reports from the contracted follow-up care providers, these reports contain only aggregate numbers (which do not correspond to the number of referrals sent in a given month) on the total number of contacts, diagnostic tests, positive diagnoses, and infants in treatment. No analysis is conducted to ensure that these numbers reflect follow-up for each infant who screened positively.
Our review of a sample of positively screened newborns from 2001 found that 150 out of 153 (98%) cases had received an initial contact to begin the follow-up process. (Note that the remaining three could not be located, but evidence of efforts to locate them were in the files.) However, the degree of follow-up care provided (i.e., if a confirmatory diagnoses was received, if counseling or instruction

was provided to the doctor, or if the infant was confirmed in care), was difficult to discern from the contractors' files without additional interpretation by staff members. In some cases, information for tracking the follow-up was not in the file and staff could not remember the details. For instance, in 19 of 65 sickle cell cases it was evident that action had been taken, but no dates were in the files to illustrate when the action was taken. Similar problems were encountered with the Emory files on metabolic and endocrine disorders. According to the Journal of Pediatrics, all follow-up actions should be documented and should include the date, who performed the action, and to whom it was directed.

During the course of this audit, the

Program initiated an electronic system

whereby MCG and Grady can report their follow-up activities on specific infants back to Program management. Discussions were also held regarding the possibility of

Best practices state that the follow-up component must ensure that the infant is located and brought into the diagnostic and management components of the system in time to ensure optimal outcome and to prevent irreversible damage.
The Journal of Pediatrics October 2000

creating such a

system for Emory. For this system to be

used effectively to ensure all positively

screened newborns receive follow-up care,

the Program must institute a mechanism

for evaluating it on an ongoing basis. In

addition to ensuring that follow-up is

conducted, Program management should

ensure that follow-up is conducted in a

timely manner and within the time frames

Genetic Newborn Screening Program

Page 15

established by law and best practices. This point is discussed further in Recommendation No. 8.
In its response to this report, the Program stated that it now plans to have a new record-keeping system in place within one year. The GPHL will send all positive screening results directly to follow-up contractors and the Program Office; and follow-up activities--including what specific actions were taken, when they occurred, and what the outcomes were-- will be recorded in electronic format in order to centralize the follow-up data. The Department plans to develop an interim electronic database for Emory. The Program's Work Group has developed a protocol on how to document actions taken regarding each infant; the protocol is in the Division of Public Health's (DPH) approval process. In addition, the Department stated that it plans to create a new position within the DPH to aid in analysis of data collected in the new system.
Recommendation No. 5
The Division of Public Health should take steps to institute internal controls throughout the Program's data transfer system in order to ensure accurate and timely transmittal of screening information to follow-up care providers. Currently, there are no controls in place to ensure that newborns' test results and demographic information are transmitted correctly, completely, or in a timely manner among the necessary Program entities. Accurate and timely transmittal of

screening information is vital to the Program's retrieval of positively screened newborns for diagnostic testing and followup care. (See the flow chart on page 4 for the necessary transfers of information among Program entities.)

A review of the Program Office's screening records revealed that the Program Manager was not receiving all of the sickle cell test results from the Georgia Public Health Laboratory (GPHL). Since the discovery, the Program has taken action to rectify the data transfer problem; however, there are still no controls in place to ensure this situation will not recur.

Additionally, the Program identified errors in its transfer of sickle cell screening results to MCG and Grady. In March 2002, the Program determined

In 2001, 38 babies who tested positively for sickle cell disorders were not referred for follow-up care. This data transmission error was identified when a physician called the MCG coordinator to ask why he had not heard from her regarding the sickle cell baby under his care. By this time, 11 of the infants were at or approaching their first birthday.

that records for

at least 38 newborns positively screened

for sickle cell disorders in 2001 had

mistakenly not been transmitted to the

follow-up care providers when they were

initially screened. The missed infants

came to the attention of the MCG sickle

cell follow-up coordinator during the

course of the audit. The coordinator was

contacted by a physician who had received

a positive screening result from the

laboratory, but no contact from MCG. The

Genetic Newborn Screening Program

Page 16

MCG follow-up office had no record of this infant's positive screening. After further investigation, it was determined that this newborn and at least 37 others (27% of all newborns reported to have screened positively for sickle cell disorders in 2001) had not been transmitted to the MCG and Grady follow-up coordinators by the Program Office. When the problem was detected in March 2002, 11 of the 38 infants were at or approaching their first birthday.

Both MCG and Grady have now taken action to locate these infants and get them into care. The immediate data transfer problem that caused these babies to be missed also has been corrected. It should also be noted, however, that until April 2001, the laboratory maintained no record of all the positive screening results sent to Emory for follow-up, so there was no record against which to check what Emory had received. These records are now maintained in hard-copy form, but no assurances or controls have been instituted to compare these files to what Emory has actually received.

In order to ensure all newborns' results are

released to follow-up personnel responsible

for notifying parents and physicians of the

disorder and ensuring a diagnostic test is

completed and treatment is begun, the

Division must institute internal controls

throughout the Program's data transfer

system. In addition, Program Management

Without a by-baby review of follow-up activities, errors go undetected by the Program.

should ensure that all data necessary to evaluate Program

activities and correct any problems detected is collected and transmitted correctly and completely among the Program entities. This point is discussed further in Recommendation No. 13.
According to the Department's response to this report, the Program Office has begun weekly comparisons of the information sent to MCG and Grady and its master list of screening results. In addition, the Department "will be working with Emory to reprogram its data reporting ... as a temporary measure [for approximately 18 months until follow-up is centralized in a new data system].... An electronic file that documents phone calls, confirmed diagnosis, initiation of treatment, dates of actions and resolution of referred results will be returned to the Genetics Program."
Recommendation No. 6
The Program should conduct regular and ongoing assessment of the Voice Response System (VRS) to ensure the system is functioning properly and reporting test results in a timely and accurate manner.
The Program's Voice Response System (VRS), through which physicians can call into the Georgia Public Health Laboratory (GPHL) and obtain test results, was created as a safety net to ensure physicians' access to the results. However, testing of the system identified numerous technical problems, as well as

Genetic Newborn Screening Program

Page 17

errors in the actual test result reports.
We tested the VRS on three different days using randomly selected specimens. As detailed below, we identified significant technical and result-reporting problems.1
Test results reported through the VRS differed from the results mailed to physicians and hospitals. For example, in one case, the mailed result indicated that the infant had a positive congenital hypothyroidism result; however, the VRS reported that the tests for congenital hypothyroidism were normal.
The VRS incorrectly identified an old sample as the most recent one on file, even though the more recent results could be accessed.
In seven cases, the VRS stated results were "not on file"; however, mailed results had already been released.
The VRS did not consistently indicate if blood samples tested were collected when the infant was under 48 hours old. Collecting specimens before this deadline affects the accuracy of results and physicians' interpretation of them, and requires repeat screening.
Results stated that the infant had screened positively, but the system did not identify the specific disorder(s) for which the infant had screened positively.
Technical problems encountered while using the VRS included the following. The system: Referred to an incorrect phone number
for obtaining further information; Disconnected the caller (without

explanation) when no birth date had been entered into the system; Did not link to a staff member when requested by the caller; and Identified multiple tests on the same infant as results for individual children.
In addition, only 7 of the 19 samples tested (approximately 37%) had a mother's SSN recorded to be used to access the results. While the mother's SSN may have been unavailable to the laboratory, doctors would have difficulty obtaining results unless they had a copy of the actual blood sample collection form that includes the lab form number. (As noted on page 5, the laboratory is in the process of phasing-in a new blood sample collection form that includes a tear-off sheet for the mother including the lab form number.)
GPHL staff were alerted to the problems encountered by the audit team during the course of the audit and have stated that efforts have been made to resolve some of the problems. These efforts should continue, and procedures for the GPHL to perform its own comprehensive check of the system on a regular basis should be implemented.
The laboratory's current method of reviewing the VRS's operation is insufficient to detect and correct errors in its technical functioning and reporting. According to GPHL staff, one specimen per week is randomly selected to test the VRS. There is no comprehensive, systematic check of the VRS's operation.

(1) Note that in each of these cases the mailed test results had already been released. The results on file at the GPHL were used for comparison to the VRS-reported results.

Genetic Newborn Screening Program

Page 18

The laboratory's current method of reviewing the VRS's operation is insufficient to detect and correct errors in its technical functioning and reporting.

If the VRS gives incorrect information for this specimen, or if the specimen is not accessible

through the system, a staff member

corrects the information and/or enters the

correct results. However, no further action

is taken to determine if other samples have

similar problems. It should be noted that

the laboratory receives an average of 1,034

blood samples for testing each day. In

addition, laboratory management has stated

that it depends on doctors that use the VRS

to notify the laboratory of any problems

with this system. However, our review

identified problems physicians could not

detect through regular usage.

obtain a screening result depends on which Program entity they contact. No standards exist outlining when verbal or written results can be released, or when the physician must be forwarded on to someone else.
For instance, if the physician calls the Program Manager's office, he or she will be given a verbal test result, but no written copy. If the physician calls the Emory follow-up office, he or she will be given a verbal test result and may be sent a faxed copy of the unofficial result report. If the physician calls the laboratory and is not the physician listed on the sample form, he or she will be told to call one of three individuals depending on the result. No results will be released to the physician by the laboratory.

Without implementing ongoing, thorough checks of the VRS's operations, the Program may continue to provide physicians with incorrect, conflicting, and confusing reports from the system. In addition, technical problems may go undetected, resulting in decreased use of the system.
In its response to this report, the Department stated that a quarterly plan to test the VRS is being developed and will be implemented in December 2002.
Recommendation No. 7
The Program should institute a standard protocol for the release of newborn's screening results. Currently, whether or not physicians can

The federal review of the Program conducted in 1999 also recognized this problem, stating that Program rules are somewhat unclear regarding reporting responsibilities. ... The Program needs to formalize the rules regarding reporting and communication in the form of Standard Operating Procedures.... In order to protect infants' medical records and ensure screening information is rapidly distributed to those who need it for the child's treatment and well-being, the Program should standardize: what information will be released (and in what format); to whom information will be released; and, what security measures should be in place before results are released.
In its response to the report, the

Genetic Newborn Screening Program

Page 19

Department stated that a protocol and policy for releasing infants' screening results has been drafted and is in the Division of Public Health's policy approval process.

Exhibit 6

Screening Process Timeframes from Birth through Initial Notification of Parent or Physician regarding Positive Screening
Results

Sickle Cell Disorders

Days 19
18

Stage
MCG and Grady Initiate Follow-up and Achieve First Contact

4 days

17

16

15

Testing Results Faxed to MCG and Grady

14

13

Metabolic and Endocrine Disorders

Days 15 14 13

1 day

Stage
Emory Initiates Follow-up and Achieves First Contact Testing Results Sent to Emory

7 days

12

12

7 days

11

11

10

10

9

9

8

GPHL Receives Sample

8

7

7

GPHL Receives Sample

6

6

6 days

5

5

6 days

4

4

3

3

2

Blood Sample Collected

2

2 days

1

1

Blood Sample Collected

1 day

0

Birth

0

Birth

Source: File Reviews

Recommendation No. 8
The Program should monitor the timeframes in which the screening and follow-up processes are completed to ensure they generally comply with the three-week time frame allowed by law. The Program currently conducts no analysis of the timeframes in which the screening and follow-up processes are completed. Such an analysis is necessary to ensure compliance with the law and to effectively evaluate the Program's activities and institute any corrective action necessary to expedite screening and followup.
We evaluated the timeframes for screening and follow-up using a sample of the newborns positively screened in 2001. Timeframes were compiled using the median number of days it took to complete each stage of the process (see Exhibit 6 on the left). Our file review determined that test results completed the screening process, through the point of initial contact with the parent or physician, within 15 days for metabolic and endocrine disorders, and within 19 days for sickle cell disease. However, the law requires that the newborns' diagnostic tests be completed within three weeks of life.1 For metabolic and endocrine disorders, the timeframes compiled leave only six days to complete the diagnostic test. Our evaluation also found the median time of blood sample collection for metabolic and endocrine disorders to be one day (24 hours). Laboratory standards state that a sample must be collected at 48 hours for accurate

(1) Note that the statute does not specify a time frame for sickle cell disorder follow-up, as it does for metabolic and endocrine disorders.

Genetic Newborn Screening Program

Page 20

testing. Time to obtain replacement specimens for unsatisfactory samples would easily place the screening timeframe outside of that allowed by law. Furthermore, the Program allows for initial samples to be collected as late as seven days after birth.
The Program should institute an analysis of timeframes to ensure that they are in compliance with the law and newborns are screened and receive necessary treatment in a timely manner. Without such ongoing review, management cannot ensure that affected infants enter care within the three week timeframe allowed by law. Nor can the Program assess what corrective action or improvements are necessary to expedite screening and follow-up. Some actions that would decrease the amount of time it takes to complete the screening process were identified by the 1999 federal program review and are outlined in the following recommendation.
The Department stated in its response that it plans to analyze individual infants' screening timeframes when it has a new data system in place in approximately 18 months. In the meantime, "a proposal is being developed for a programmer to develop an interim program to document individual follow-up by Emory." The Program also has drafted a new protocol regarding what data should be recorded in order to document Program activities. The protocol is currently in the Division of Public Health's approval process.
Genetic Newborn Screening Program

Recommendation No. 9
The Program should consider instituting recommendations provided by the federal review conducted in 1999. The 1999 federal review of the Program provided the Division with extensive direction on how its activities could be improved to provide more efficient and effective services, appropriately target its testing, and improve oversight of overall Program operations. Some of the recommendations have already been implemented by the Program; however, as discussed below, there are several areas that have not been sufficiently addressed by the Program, and are important to the improvement of overall operations.

The Program should consider no longer

testing unsatisfactory samples received

by the Georgia Public Health

Laboratory (GPHL). According to the

federal review, once a sample has been

declared unsatisfactory for analysis

according to a prescribed protocol,

then to perform a laboratory analysis

increases the legal exposure of the

laboratory. The GPHL currently

completes all screening tests on

samples it receives and deems

unsatisfactory for accurate results.

According to

laboratory staff, the samples are tested in case a replacement is

Once a sample has been declared unsatisfactory for analysis according to a

never received, even though the accuracy of the tests cannot be ensured.

prescribed protocol, then to perform a laboratory analysis increases the legal exposure of the laboratory.

Results on these samples are not

1999 federal review of the Program

reported unless

Page 21

one or more of the tests is positive. It should be noted that if the Program institutes procedures to ensure it receives an adequate sample to replace the unsatisfactory one (see Recommendation No. 3), the need to test unsatisfactory samples received may decline. In addition, education of hospitals regarding their unsatisfactory sample submissions may result in a decrease in the number of unsatisfactory samples received.
The Department indicated in its response that although there will be no immediate changes to the laboratory's testing policy, the Program's Work Group "is reviewing and considering reasons for unsatisfactory specimens. [The Department expects] the number of unsatisfactory specimens to dramatically decrease as a result of Hospital Reports and targeted education and consultation, and will monitor these trends. [The Department will] determine the value of continuing to test unsatisfactory specimens after planned strategies and actions in the [Program's] work plan are implemented."
The Program should evaluate the benefits to and possibility of expanding the laboratory's hours of operation. According to the federal review, most states have a five and one-half day week [testing schedule] with results completed on Saturday mornings for some analyses ... [and] individuals rotating responsibility for follow-up on Saturday or over ... long weekends. As operations are currently conducted, the GPHL is open five days per week, but because some tests run overnight, only runs tests on four of those five days. The laboratory also takes all state
Genetic Newborn Screening Program

holidays, which further delays testing, especially if the holiday falls in the middle of the week. One method the follow-up contractors have used is to establish an emergency contact system for parents and physicians to obtain information or refer a newborn; the Program does not currently have such a mechanism.
In its response to the report, the DHR stated that the GPHL is considering the option of expanded hours of operation as more tests are added to the state's screening program.
The Program should conduct assessments to determine if physicians are aware of and using the Voice Response System (VRS). The federal review stated that the VRS is an important and timely addition to the newborn screening program ... [however] some major users interviewed indicated that they were unsure if they had been notified and issued an PIN number, and all agreed that any information received had been limited. The Program provided us with letters sent in September 1998 and October 2001 to notify physicians about the VRS, how to register, and how the system can be used. The VRS is also explained on the Program's web site, and may have been mentioned at the educational seminars conducted in 2000 and 2002. However, as stated three years ago in the federal review, usage monitoring reports are available and their routine use is recommended to monitor the goal of increasing system usage. In addition, additional publicity may be needed.
The Program should be commended for adding a toll free number to the VRS,
Page 22

as recommended in 1999, and for moving forward with the 1999 review recommendation to add faxing capabilities to the system in order to meet customer needs for hard copies of results. In March 2002, the laboratory obtained an estimate of $8,700 to add faxing capabilities to the current computer system (which is expected to change within the next two years).
In its response to the report, the DHR stated that it will conduct an assessment of the VRS's current users by February 2003. In addition, the Program is preparing a survey for medical providers to determine their awareness, use, and satisfaction in using the system. The survey will also request suggestions on improvements to the VRS. Survey results will be incorporated into a revised VRS when the GPHL changes computer systems [in approximately two years]. The Department also noted that installation of faxing capabilities is scheduled to occur in late November 2002.
The Program should evaluate new technology in light of the costs associated with the new technology, as well as the overall benefit to the Program. As indicated in the federal review, the theoretical increased accuracy and expanded capabilities available using more expensive techniques available to newborn screening laboratories may not be necessary for a screening program. Examples of these technologies are tandem mass spectrometry (MS/MS), which allows for testing for over 30 disorders, and DNA testing. In addition, the ability to test for more disorders may not mean that screening for these disorders should occur. New
Genetic Newborn Screening Program

screening tests should be evaluated in light of set criteria for screening established by the Program. (See Recommendation No. 10.)
Improved technologies can also facilitate increased accuracy in blood specimens collected in younger newborns. Currently, Georgia's laboratory procedures designate blood samples collected when the baby is under 48 hours old as insufficient to detect all disorders. In these cases a replacement sample is required for complete and accurate screening. With improved technologies, the laboratory could consider moving to a 24 hour after birth collection time standard, as some states have done. In calendar year 2001, the laboratory reported receiving 27,321 blood specimens collected at under 48 hours of age (13.5% of the 202,577 samples received that year). It should be noted that newborn screening experts differ in their support of a 24 hour versus 48 hour optimum screening time.
The Program should educate the hospitals' Advisory Committee representative about the critical nature of hospitals' input on Program operations. As noted by the federal review, it is important for the Program to have the input and active participation of hospitals in the newborn screening program. There is currently a place on the Advisory Committee for a representative from the Georgia Hospital Association; however, this representative has not attended any Committee meetings in over two years. The Program should examine ways to encourage active hospital representative participation.
Page 23

In its response to this report, the DHR noted that it has recently solicited the Georgia Hospital Association's (GHA) input regarding blood specimen quality reports mailed to hospitals (see Recommendation No. 3) and hospital training activities. The Department also stated that it is soliciting the GHA's active participation on the Program's Advisory Committee.
The Division should ensure that a central authority for the Genetic Newborn Screening Program has been clearly identified for all Program entities. As stated by the 1999 federal Program review, it is essential that the central authority be clearly identified ... have decision-making authority, and be able to take actions that will improve the overall function of the Program. During our review, we found that various Program staff were making policy decisions about such things as whether or not to test unsatisfactory samples, what follow-up care information would be released and by whom, and how replacement samples for specimens deemed unsatisfactory for testing would be requested.
By addressing these recommendations from the federal review, the Program could significantly improve its efficiency and better target its efforts to screen infants for relevant disorders. It could also better the Program's external educational efforts and internal communications and management.
Recommendation No. 10
The Program should establish formal criteria to use in determining the disorders for which the state will screen.

Currently, the disorders for which the Program screens are all outlined in state statute. The statute also allows for the addition of metabolic disorders as may be determined in the future to cause mental retardation if undiagnosed and untreated. However, no criteria have been established for the Program to evaluate the continued or future appropriateness any of these disorders for the newborn screening program.

Best practices outline criteria for how to

evaluate disorders for inclusion in newborn

screening programs. Among the reasons a

disorder should or

should not be included are its incidence rate within the population, the cost for conducting the screening, and the availability of treatment if an

In order to provide the legislature and its advisory groups with information necessary for future policy decisions, the Program should conduct ongoing evaluation of the appropriateness of including disorders in the screening program.

infant is diagnosed

with the disorder. In addition, improved

technology can make testing for new

disorders feasible for a screening program.

In these cases, best practices state that the

population should be tested through a pilot

study to see if the incidence rates warrant

adding the test.

The Program is currently considering two expansions to its screening battery-- biotinidase deficiency and MCAD (medium chain acyl-CoA dehydrogenase deficiency). However, without established criteria regarding the disorders for which the state will screen, the appropriateness of

Genetic Newborn Screening Program

Page 24

these disorders for the Program cannot be determined. The Program has drafted and submitted for approval revised Rules that lay out screening criteria for the Program. These criteria are consistent with Best Practices and the recommendations of the federal review of the Program conducted in 1999. However, the new Rules also include the addition of biotinidase deficiency and MCAD to the state's screening tests; but no study has been completed to determine if these disorders meet the new criteria.

The 1999 federal review of the Program stated that all additions and deletions of tests [to Georgia's Screening Program] should be considered on the basis of incidence, treatment, cost, and overall benefit, with minimal political consideration.

In order to provide the legislature and DHR advisory groups with the information necessary for future screening policy decisions,

the Program

should conduct ongoing evaluation--based

on clear criteria--of the continued

appropriateness and necessity of disorders

for the screening program.

In its response to this report, the DHR stated that the addition of biotinidase deficiency and MCAD to the state's screening "is based on experiences of other states and the recommendation of [the] national March of Dimes."

Recommendation No. 11
The Program should review hospitals' screening practices and target training and educational efforts to those

institutions with the greatest need. Currently, the Program does not monitor hospitals or provide feedback on their newborn screening-related activities. The Program has no assurances that: hospitals are providing parents with information regarding their right to refuse screening or on the potential need for re-screening; blood samples are collected in the correct manner; accurate and complete demographic data is provided on the screening forms; and samples are appropriately transported to the laboratory. As a result, the Program cannot ensure that hospitals are effectively screening all newborns. Additionally, targeting training and educational efforts at hospitals identified through analysis of which hospitals are not submitting screening specimens or are submitting a large number of unsatisfactory specimens (see Recommendations Nos. 2 and 3), could improve Program operations.

Hospitals currently define their own

procedures for handling religious

refusals and parent notification

regarding re-screening; and there is no

review by the Program to determine if

they are in compliance with these stated

requirements. State statute states that

any infant whose parents object [to

screening] on the grounds that such

tests and

treatment conflict with their religious tenents and practices shall be

The Program should analyze hospital compliance to identify problems and target educational activities to hospitals with more serious difficulties.
1999 federal review of the Program

Genetic Newborn Screening Program

Page 25

excluded from the testing requirement. According to the Program's screening manual: If a parent objects to testing based on religious grounds, a hospital official is to inform the parent of the consequences of refusal (possible infant death or retardation) and require the parent to complete a statement indicating their declination of newborn screening for religious reasons. This signed refusal should be retained in the record.... In addition, infants who are released from the hospital at under 48 hours of age are to have another blood specimen collected for testing within one week of life. This is to ensure accurate screening. In such cases, the Program requires that the hospital administrator or a designated representative ... provide written notice to the parents, guardian or legally responsible person that the infant must be tested again prior to one week of age. The Program should continue its efforts to educate hospitals on the appropriate way to carry out these responsibilities, and should review hospitals' implementation. It also should be noted that were the Program to have knowledge of the number of documented religious exemptions through review of hospital records, required reporting of religious exemptions, or another means, it could better evaluate its screening rate.
In its response to the report, the Department stated that it has developed a protocol for hospitals to follow in conducting their newborn screeningrelated activities. This protocol is discussed with those hospitals that
Genetic Newborn Screening Program

request or are targeted for training.
According to Program staff, the number of unsatisfactory samples markedly declined after training sessions were offered in 2000. Such training was scheduled (and is partially completed) again for 2002. The Program should evaluate how often this training is needed and for whom it should be given, taking into account high turnover in hospital staff. As noted in Recommendation No. 3, the Program has plans to regularly notify hospitals of their unsatisfactory sample rate, the reasons their samples were unsatisfactory, and how to improve them. The first of these hospital reports were mailed in August 2002.
Follow-up personnel have indicated that inaccurate and missing demographic data can hinder their retrieval of positively screened infants. The Program plans to include information on hospitals' provision of this information in its hospital reports, as well. Notifying hospitals about their performance level in providing infants' information may limit this problem.
The Department noted in its response to this report that hospitals have reacted positively to the Program's feedback on their performance. Additionally, in response to the audit's recommendation, the Program is planning to further analyze hospitals' data and target training to those hospitals experiencing the greatest problems.
Page 26

In addition, as recommended by the 1999 federal review of the Program, hospital personnel need to be educated about the liability associated with sending groups of blood specimens in batches. This practice results in many newborns' specimens being delayed in their shipment to the laboratory, potentially furthering the duration of a disorder and possibly resulting in permanent damage.
As recommended in the 1999 federal review, the Program should analyze hospital compliance to identify problems and target educational activities to hospitals with more serious difficulties. The Program's review should include: evaluation of hospitals' blood specimen collection and transport, as well as correct and consistent recording of demographic data on the collection form, and hospitals' provision of parents' right of religious refusal. Reviewing hospitals' activities and targeting training and education appropriately could reduce the number of unsatisfactory samples received and provide information for evaluating screening rates. Additionally, such action would help ensure that parents are afforded the rights provided them in the law.
The DHR stated in its response that it plans to contract "with the March of Dimes to work with the Newborn Screening Advisory Committee to develop educational materials [regarding the availability of] expanded screening [tests], including an interactive CD Rom for professionals, laminated fact sheets, posters and an Action Plan to disseminate materials to OB/GYN, pediatricians and family
Genetic Newborn Screening Program

practitioners." See the related Informational Finding No. 1, p. 31.
Recommendation No. 12
The Division should determine the costs associated with each component of the Program, evaluate the reasonableness of these expenses, and consider what funding sources are available.
The Department has conducted no evaluation of the Program's current costs, future financial needs, or available funding sources. Program management was not able to provide a cost per screening estimate, nor a total cost estimate for maintaining the Program. Neither has the Department done any assessment of the contracted follow-up care providers' actual expenditures or financial needs. In its Guidelines for Newborn Screening, The Journal of Pediatrics states that evaluation of the overall newborn screening system should include cost per patient per period (month, year, or lifetime), cost of each component service...costs for operating the entire system, the cost benefit to patients, families, and society...[and] analyses of costs and outcomes. And the 1999 federal review of the Program stated that the resources dedicated to the Program must undergo continued reassessment to assure optimal functioning....
In order to determine the actual costs to run the Program, we had to consult financial personnel at each of the individual contracted entities, in the Georgia Public
Page 27

Health Laboratory, and in the Family Health Branch of the Division of Public Health. The contracted follow-up care providers currently submit an annual budget justification for the Department's approval. Program management reported, however, that no analysis has been done by the Program to assess the contractors' actual necessary expenditures. In addition, the contractors submit monthly expense reports for reimbursement; but no analysis is done to correlate the budget justification to the expenditures reported.
Some money provided to the contracted follow-up care providers is dedicated to treatment. However, no analysis has been done by the Program to determine the financial need for this state support given that alternative sources of funds, such as insurance, might cover some of these costs. The Program should include in an analysis of the Program's actual costs an evaluation of what other funding sources may be available.
In its response to the report, the DHR stated that it has used the cost information complied in the audit to evaluate its current expenses. The DHR has worked with the Department of Community Health to obtain an additional $2.4 million in Medicaid matching funds, and is continuing to seek out other funding sources. The Department plans to use the additional funds to add positions to the GPHL, purchase laboratory equipment, add two Program staff, and provide additional funding for follow-up at Emory.

Recommendation No. 13
Program Management should ensure that each component of the Program is operating effectively by ensuring they collect and report data necessary for ongoing assessment of the Program's effectiveness. As stated by the 1999 federal review of the Program, it is critical that productivity indices be developed for individuals at all levels in the Program as a means of evaluating achievement of optimal performance. These indices should be periodically monitored and corrective actions taken when not achieved.
In conducting our analysis of Program effectiveness, we found that data was often not compatible or available. For instance, there is no central, correct, and accessible record of all the positive screenings detected by the Program. Follow-up contractors' records were kept in various formats on-site, and in all cases required interpretation by the follow-up coordinators. In addition, when comparing birth and screening records, we found that different information is recorded in the two databases and different policies are used on how to record the data. For instance, the manner in which hyphenated last names are recorded differs in the two records, as does the way in which race is denoted. As a result, it was difficult to compare the two databases.
In order to effectively evaluate the Program, management should define what information is necessary for sufficient Program analysis, and ensure that the laboratory and contracted follow-up care

Genetic Newborn Screening Program

Page 28

providers are recording the data in such a way that it can be used for this analysis.
In its response to the report, the DHR stated that information regarding screening results and follow-up activities will be centralized into a uniform database in approximately 18 months. "The data system is being designed to facilitate the assessment and evaluation of the Program." A protocol on how to document actions taken regarding each infant is currently in the Division of Public Health's approval process. In addition, discussions regarding the consistency of data in the birth and screening records have been held among Vital Records, the GPHL, and Program staff.
Recommendation No. 14
The Program should systematize coordination with other health care programs within the Division of Public Health to ensure a seamless transition of affected infants into care. The Division has many resources available to infants and families affected by these genetic disorders, including services to provide on-going case management and supply medical referrals, treatment, and equipment. However, there is currently no formalized mechanism by which infants detected with disorders through the Genetic Newborn Screening Program are referred into these programs.

the Department's adherence to statutory requirements that along with instituting the Program, it establish a state-wide network to provide counseling regarding genetically caused disorders and for initiating and continuing therapy.
It should be noted that since this recommendation was brought to the attention of Program staff, coordination with other Division of Public Health programs has been added to the Program's future work plan.
In its response to this report, the Department explained that under its new system, each positively diagnosed infant will be referred to the Children 1st Program, which serves as a clearinghouse for referrals into all other Division of Public Health programs. Children 1st will conduct a home visit, refer all infants into Children with Special Needs programs, and determine what other referrals are appropriate.

Developing such a system would maximize the effectiveness of this Program and others within the Division of Public Health. In addition, such coordination would support

Genetic Newborn Screening Program

Page 29

Recommendation No. 1
Consideration should be given to revising existing Genetic Newborn Screening statutes to fully reflect the activities of the Program and clarify the Department's responsibilities regarding future changes to the Program. Although current statute does outline many of the facets of the Genetic Newborn Screening Program and the system of genetics services overseen by the Program, it does not completely reflect the Program's current activities. Nor does current statute necessarily provide the Department with the latitude it appears to intend. These discrepancies could be rectified through the revisions noted below.
Currently, statutes outlining the process for screening, retrieval, and diagnosis only provide for this system of followup for the metabolic and endocrine disorders, even though the Program provides the same type of follow-up care for sickle cell disorders. To more clearly reflect Program activities, the statute could be amended to include such sickle cell disorder follow-up. It should be noted that sickle cell disorders are considered by Best Practices and the Program to require follow-up of the same type provided for metabolic and endocrine disorders, even though there is less urgency associated with them.

Department with the latitude to expand screening as deemed necessary, consideration should be given to revising the current wording to include all metabolic and endocrine disorders, as some of the current disorders are categorized as endocrine disorders. Additional consideration may be given to expanding the requirements for adding a new disorder to include those that may result in death, as well as mental retardation.
State statute currently requires parental notification about infants' positive screening for only two of the eight disorders. Given the severity of all the disorders screened for by the Program, the Legislature may consider revising this section of the code to include all disorders.
Such revisions to the statute would more clearly define the parameters of future screening and follow-up efforts and more accurately reflect in the statute the Program's current activities.

Current statute limits the disorders that may be added to the screening program to metabolic disorders as may be determined in the future to cause mental retardation if undiagnosed and untreated. If the intent is to provide the

Genetic Newborn Screening Program

Page 30

Finding No. 1
At least one other state has legislatively mandated that pediatricians and obstetricians notify all newborns' parents about the availability of genetic disorder screenings not currently provided by the state. Georgia currently screens for eight metabolic, endocrine, and sickle cell blood disorders. Technology exists to test for over 700 genetic disorders; but because not all disorders or tests are suitable for a statewide screening program, each state has determined the specific disorders for which it will manditorily screen. In Mississippi, however, the legislature determined in 2001 that mothers should be notified of the other disorders for which their newborns can be screened. As stated in the statute, any clinic, hospital, physician or health care provider providing prenatal care to a pregnant woman ... [and] the physician or other health care provider attending a newborn child shall notify the parents that there are newborn screening tests available ... in addition to the tests that are required by the state, and shall provide to the parents the most recent information developed by the State Department of Health regarding these tests. Currently, there is no standard literature or other information provided to Georgia parents regarding additional genetic tests available outside of the state's Genetic Newborn Screening Program.
The DHR stated in its response to the report that "a brochure to make providers and parents aware of supplemental

screening tests is being developed and will be distributed by the March of Dimes and the Georgia Chapter [of the] American Academy of Pediatrics."

Finding No. 2

Current state statute regarding the

privacy of genetic information does not

apply to the Genetic Newborn Screening

Program tests.

Georgia's genetic testing statute excludes from its definition of genetic testing chemical [and]

Georgia's statute regarding the privacy of genetic information excludes from its definition of genetic testing chemical [and] blood analysis such as that used by the Program.

blood analysis such as that used by the

Program. However, as stated in the federal

review of the Program in 1999, and noted

by laboratory staff, newborn screening is

moving towards the use of DNA testing to

identify disorders. If that were to occur, the

screening program would come under the

purview of the state's current genetic

information privacy statutes, as DNA test

results are included in the definition of

those protected.

Finding No. 3
Effective April 2003, the Program's processes for managing and protecting newborns' screening and follow-up information will be subject to the federal Health Insurance Portability and Accountability Act (HIPAA) of 1996. The federal Health Insurance Portability and Accountability Act (HIPAA) of 1996

Genetic Newborn Screening Program

Page 31

requires that all individually identifiable health information remain private and secure. The Act also creates standards regulating the electronic transmission of health information, such as that handled by the Program. Federal Rule requires that the Department ensure the privacy of all health information by April 2003. The deadline

for compliance with the electronic transmission standards is October 2003.
Although the audit did not address the Program's compliance with HIPAA, the Department acknowledged in its response to the report that HIPAA regulations will affect future information-exchange systems.

Genetic Newborn Screening Program

Page 32

Appendix A Disorders for which Georgia's Genetic Newborn Screening Program Tests

Genetic Newborn Screening Program

Disorder

Time of Onset and Impact

Life-Long Treatment

Symptoms

Congenital Adrenal Hyperplasia (CAH): Lack of an enzyme used to produce hormones necessary for regulating proper growth,development, metabolism, and fluid balances in the body. Galactosemia: An inability to break down sugars from milk and other products due to the absence of a particular enzyme.
Homocystinuria: An inability to change one amino acid into another, resulting in the toxic build-up of the amino acid methionine in the blood.
Congenital Hypothyroidism: The inability of the thyroid gland to produce sufficient amounts of the growth-related hormone thyroxine.
M aple Syrup Urine Disease (M SUD): An inability to breakdown the amino acids found in protein due to an enzyme deficiency. This ultimately prevents the baby from being able to break down proteins in food and milk.
Phenylketonuria (PKU): An inability to break down proteins in food and milk due to the absence of a liver enzyme necessary for the digestion of proteins.
Sickle Cell Blood Disorders: An abnormal version of the blood protein hemoglobin causes red blood cells to change shape. As a result, the blood cells clog blood vessels, preventing oxygen transport to other parts of the body. There are varying types and severities of this disorder. Tyrosinemia: An inability to metabolize the amino acid tyrosine, present in most proteins, causing a toxic buildup in the blood.
Source: Professional Literature

Most cases are apparent within the first week of life. CAH can result in ambiguous genitalia in females, short stature, severe dehydration, and a failure to grow at the normal rate. CAH will lead to rapid death of an untreated child. Infants may appear normal at birth, but will often present symptoms within the first week of life. Untreated infants will normally die within the first week of life. Those treated late are at a high risk for mental retardation, cataracts, spasticity, sepsis (blood infection), and disease and/or enlargement of the liver.
Infants are usually symptom free for the first few months of life. Failure to treat can result in developmental and neurological problems. Additional impacts include glaucoma, blood clots, scoliosis, osteoporosis, mental retardation, systemic organ failure, and dislocated ocular lenses.
The disorder is present at birth, but untreated infants may appear normal up to three months of age, by which time some brain damage will likely have occurred. Non-treatment will lead to growth failure and mental retardation.
Infants are very ill at birth and are usually already hospitalized for other, yet related, problems. Untreated infants will die within the first three months of life.
The impact is seen within the first 24 hours of life. If treatment is not begun immediately (within the first 2-3 weeks of life), cognitive ability will be impaired and ultimately lead to severe mental retardation and a decreased life-span. Most individuals not treated early will achieve an I.Q. of less than 50. Other impacts may include autistic-like disorders and growth failure.
The impact is immediate and may include any of the following: anemia, chronic tissue damage, severe pain, a lethal decrease in the production of hemoglobin (part of the cell that carries oxygen), potentially fatal sepsis (blood infections), and strokes.
W ithout treatment the disease will be fatal within the first year of life due to acute liver disease and possible hemorrhage. Other impacts include liver failure, neurological crises, progressive paralysis, hypertension, and corneal erosions.

* Hormone and cortisone treatments * Constant follow-up care by a pediatric endocrinologist
* Elimination of milk and milk products from the diet * Patients who are treated promptly usually have normal survival and improved physical and neurological outc om es . * Specific diet modifications * Large doses of vitamin B12 * Supplements of folic acid
* Hormone therapy in pill form will lead to normal growth and d evelop m en t
* Dietary restrictions requiring a complicated, low-protein formula * Intensive monitoring by a medical pr of es s i on al
* Dietary restrictions requiring a special diet (including special infant formula available through the P r ogr am ) * Counseling for adult females on the impacts PKU can have on the fetus during pregnancy if the diet is not maintained * Antibiotic therapy until age 5 * Pain management * Blood transfusions * O xygen treatments * Surgery
* Dietary restrictions * Even with treatment, the condition may necessitate a liver transplant to prevent liver cancer and its spread to other parts of the body.

Vomiting, weakness, shock, and irregular heartbeats
Jaundice, vomiting, feeding intolerance, diarrhea, and lethargy
Developmental and neurological problems, mental retardation, seizures, psychiatric disturbances, flushness of cheeks, a tall thin build with long limbs, n ears ig ht ed n es s Jaundice, constipation, lethargy, poor muscle tone, feeding problems, a large tongue, mottled and dry skin, deafness, an enlarged belly, and neurological ab n or m alities Odor of maple syrup in the urine, a high pitched cry, irritability, spasticity (a jerkiness of the limbs), seizures, poor feeding habits, coma, mental retardation, and l et h ar g y Autistic-like behavior, skin rashes, tremors, spasticity, delayed social and mental development, seizures, and a peculiar body odor
Pain in the abdomen, joints, and bones, weakness and fatigue, delayed growth, infections, jaundice, and fever
Failure to physically develop at a normal rate, vomiting, diarrhea, a cabbage-like odor, severe light sensitivity, and progressive p ar alys is

Page 33

For more information or to obtain additional copies of this report, please contact Paul E. Bernard, Director,
Performance Audit Operations Division, at 404.657.5220.

Genetic Newborn Screening Program

Page 34