Background. Newborn screening saves lives, but psychosocial complications after genetic screening have led to doubts about expanding programs. Because complications have been blamed on ineffective communication of results, a population-scale system to ensure communication quality may improve outcomes. The objective of this study was to develop and evaluate a method to assess the content of communication after newborn genetic screening.
Methods. We abstracted content data and calculated quantitative scores for 3 communication quality indicators (key content, early placement of good news, and excessive background content) for 59 transcribed conversations between pediatric residents and simulated parents of an “infant” who was found via newborn screening to carry either cystic fibrosis or sickle cell hemoglobinopathy.
Results. Only 8.5% of transcripts contained the key content items that were thought to be necessary for parental understanding; 27.1% included reassuring news about carrier status within the first 10% of content. Scores for 3 quality indicators fell in the low performance range in 35.6%, 30.5%, and 27.1% of transcripts, respectively. The most common topic was background about the disease (22% of content statements) even though the infant did not have the disease. Surprisingly, 50% of sickle trait transcripts included counseling about a possible risk for sudden death.
Conclusions. Assessment of the content domain of communication quality identified some high-quality communication interspersed with many missed opportunities. If integrated into newborn screening, our method may help to alleviate some of society's ethical concerns about benefit and risk after newborn and other genetic screening.
Every day, tens of thousands of US newborns are screened for serious genetic and metabolic diseases.1 As many parents learn, however, not all “positive” screening results mean that disease is present. This is consistent with Bayes' theorem, which suggests that most positive screens for low-incidence diseases will turn out to be falsely positive. Indeed, false-positive results may be >50 times as common as true-positive results.2 Many false-positive tests identify infants with benign biochemical abnormalities or heterozygous “carriers” for sickle cell hemoglobinopathy (SCH) or cystic fibrosis (CF). Communicating these results clearly to families is important because historical evidence3 and ethical concerns4,5 suggest that misunderstanding in carrier families can lead to psychosocial complications.6–9 Families of infants with diseases are often educated by genetic counselors,10 but the main responsibility for initial counseling of carrier families lies with primary care clinicians,11 especially because the supply of genetic counselors is limited enough that most heterozygous families will not be able to gain access to them.12 Unfortunately, primary clinicians' communication after newborn screening has been criticized by families13 and public health officials.10 Because communication training methods are expensive and have rarely been used on a population scale, we became interested in adapting to communication several quality assessment methods with track records for improving clinician behavior.
This study is part of a larger, multistage project to assess and improve the “quality” of communication by primary care clinicians after newborn genetic screening. The methods were designed for use in the community, but residents were ideal for this first process-centered stage given that many are near the peak of their knowledge about CF and SCH. Because “communication quality” is difficult to define precisely, we adapted the “quality indicator” strategy from biomedical quality improvement. Indicators are explicitly defined, quantitatively reliable measures of beneficial behaviors that are linked with positive outcomes and can be tracked and intervened on for improvement; each indicator represents a small part of total “quality.” For the larger project, a list of >20 communication quality indicators was produced from a detailed review of the literature.14–20 This article presents results from the 3 indicators that deal with content, ie, the subject matter of the communication (“what was said”). We chose to focus on CF and SCH screening because they are the only widely used newborn screening tests that identify genetic carriers and because most US newborn screening programs notify clinicians and/or parents of potential carrier status even though they are intended to screen for disease.1 Although the implications of a positive screening test result are different for CF versus SCH, it was important to us to have a single, adaptable method that can be used for many different disease screening programs. A reliable method to measure and compare communication processes for 2 different diseases should be a significant step toward society's need to optimize communication outcomes after genetic screening.
To assess the content domain of communication quality after newborn genetic screening, we used an explicit-criteria procedure to abstract transcripts of conversations between resident pediatricians and standardized parents of a fictitious “infant” with a positive newborn screening test. Methods and materials were approved by institutional review boards at Yale, Northwestern, and the Medical College of Wisconsin.
Participants and Setting
The participants all were Yale pediatric residents who were beyond their first year and scheduled in groups of 4 to attend a workshop on communicating heterozygous results of newborn screening tests for CF and SCH. Each workshop consisted of (1) a 10-minute review of newborn screening, CF, SCH, and autosomal recessive inheritance (but excluded any advice about how to discuss results), (2) 2 interviews with 1 standardized parent each as described below, and (3) an interactive session on brief communication of newborn screening results. The workshops were part of the official curriculum, but residents gave informed consent and were free to decline the use of their tapes in research.
The counseling task was to communicate the results of the infant's test to the standardized parent. A handout provided the resident with the test result and some social background data but did not prompt how to discuss the result. The infant's mother and father were portrayed in the scenario as both being adopted so that the session could focus more on the clinicians' risk communication than on taking an extended family history. In the SCH carrier scenarios, the handout reported a screening result of hemoglobin F, A, and S, a result that had been presented in the review session as definitely indicating that an infant is an SCH carrier. In the “likely CF carrier” scenarios, the handout reported an elevated screening immunoreactive trypsinogen level and the presence of 1 ΔF508 mutation with no multiallele follow-up screening. The review sessions had presented such a result as suggesting that the infant was probably a carrier but still has up to a 5% to 10% chance of having the disease caused by an undetected allele.21,22
During the year, 6 different standardized parents worked on the project; all were women and chosen to depict plausibly the age and the ethnicity of a mother of an infant with CF or SCH. Each participant was audiotaped in a randomized, double-crossover manner with 2 of 4 possible standardized parent encounters varied by the 2 diseases and by 2 script types. The 2 script types were included for a separate validation study of the new script type, the Brief Standardized Communication Assessment (BSCA). The BSCA is a streamlined version of a standardized patient script that simplifies role instructions, requires less training, discourages improvisation, and encourages clinician counseling to continue as long as possible with nonleading questions, such as, “Is there anything else I should know?” The comparison script type was designed to resemble the methods used in most educational settings, which seek to make the encounter realistic by asking the patient to use acting skills and improvise a 2-way dialogue.23 For helping comparison, all standardized parents in both scripts were coached not to appear anxious and to avoid leading questions of the sort that “helpful” standardized parents often ask. They were also asked to pretend to understand whatever the resident said, helping us to focus on information provided rather than the resident's ability to reclarify (a critical skill but not the subject of this study).
We taped 32 residents in 2 interviews apiece. The taping equipment failed in at least part of 5 interviews, so the final sample consisted of the remaining 29 residents with 2 transcripts and 1 resident with a single transcript (total N = 59 transcripts). The tapes were transcribed verbatim and proofread for accuracy by a board-certified pediatrician (M.F.). To guide abstraction, we used a sentence diagramming process to divide the transcripts into individual concepts or ideas (statements), each with 1 subject and 1 predicate. For quality indicator 2 a “percent-statement” number was calculated by dividing the raw statement number by the total number of statements in that interview. For example, statement 20 in a 200-statement interview would be referred to as a percent-statement number of 10%, or “10% of the way into the interview.”
Our overall approach was similar to content analysis24 but used a quality-improvement-style, explicit-criteria data dictionary derived from reviews of the literature and discussions with expert clinicians. The abstraction procedure was adapted from explicit-criteria methods used in medical record review,25 except that abstractors reviewed transcripts instead of medical records. Each transcript was abstracted duplicatively by 2 authors (M.F. and A.L.), reading statement by statement. The dictionary was organized hierarchically under 9 separate “topics” and finalized via the constant comparison method26 during the first 6 abstractions. The final dictionary included 503 separate content codes nested by specificity. For example, code 1.1.3 (“screening CF carrier positive”) was more specific than 1.1 (“screening test positive”), and both were grouped under the topic “screening test” (code 1.0). When a specific code was not available for a statement's meaning, the abstractor applied the next less specific code in the hierarchical data dictionary. Up to 3 codes could be applied for each statement. Discrepancies were automatically resolved between abstractors either by including both abstractions (up to 3 codes) or by substituting a less specific code from the hierarchical dictionary. Every third transcript was discussed after abstraction to ensure quality control and consistency, following the suggestion by Feinstein.27
Statistical analysis was done using Excel (Microsoft Corp, Redmond, WA) and JMP software (SAS Institute, Cary, NC). Inter-abstractor reliability was calculated before merging or consensus using an adapted version of Cohen's method24 for “perfect” and “graduated” agreement. Agreement was rated as “perfect” when both reviewers coded the statement in exactly the same way. Agreement was rated as “graduated” and worth half a point either when reviewers' single codings differed by 1 level of specificity or when a statement was assigned >1 code but at least 1 of the codes was exactly the same for both reviewers. For the indicator grade-level scores, we assessed consistency between participants by the Cohen method with a tolerance of 1 grade-letter difference. Other data were analyzed using 1-way analysis of variance for continuous responses for categorical variables and the χ2 test for grouped categorical responses. Statistical power was >90% for individual dichotomous comparisons by gender, year in residency, and SCH carrier versus likely CF carrier.
To help categorize performance in the first 2 indicators, scores were grouped ordinally with the familiar letter grades of A (outstanding), B (good), C (mediocre), D (inferior), or F (failure). Indicator data then could be compared using the familiar grade point average (GPA) system with an A worth 4 points, etc. Indicator data with a C or lower grade were said to qualify the transcript for a hypothetical future intervention, following the quality improvement strategy of targeting interventions to clinicians who are most in need of improvement.
Quality indicator 1, “key content,” was the number included from a short list of “must have” concepts that are critical for understanding. Unlike a clinical area such as prostate cancer screening,28 newborn screening has no officially recognized list of key content items to use as a quality benchmark. We therefore derived the items from a literature review, discussions with experts, and consideration of time limits in primary care. Because a fetal-adult-sickle (FAS) hemoglobin result is specific enough to forego additional testing, 1 key item (“more testing is needed”) was calculated only for the likely CF carrier scenarios. We assumed that ideal counseling would contain as many key content items as possible (7 for likely CF carrier, 6 for SCH carrier), so we set an A to mean that all or all but 1 of the possible items were included. We defined the C or lower grades to mean 2 or 3 items or lower. The B definition was varied to account for the different number of key items in the likely CF carrier scenarios versus the SCH carrier scenarios.
Quality indicator 2, “early placement of good news” (eg, the infant is “probably” healthy) was meant to deal with the problem that although many pediatricians intuitively inform parents about carrier status before describing its implications, parents who hear a statement such as, “Your infant's screening test was positive,” are often anxious or distracted, enough to reduce comprehension of subsequent content. Although it is inappropriate to lie or provide premature reassurance to parents,29 a good middle-ground approach is to open counseling with the best “good news” possible under the circumstances. This approach can be a good way to engage counselees, reduce anxiety, and in some cases maximize counseling effectiveness.30 “Good news” was counted for any of 39 specific codes from the data dictionary, including statements that the infant is healthy or normal or that the infant is a carrier. The relative location of the first item within the interview was determined using the percent-statement number described earlier. Letter-grade definitions assumed that optimal communication would include the earliest possible good news.
Quality indicator 3, “background to individualized content ratio,” was a negative dichotomous indicator meant to operationalize the common problem of physicians who slip into a “canned lecture” about biological and other background details. Because excessive background content in counseling after carrier screening may lead parents to believe that their infant has the disease itself,31 we assumed that risk communication should include (at a minimum) more individualized than background content. Thus, the indicator calculated a ratio of the number of background statements to the number of statements about the standardized parent, infant, or relatives (topic codes 6.0 through 9.0 and 1.0 through 5.0, respectively); a ratio >1:1 suggested a need for improvement. For GPA comparisons, a ratio of 1 or more (the “excessive” score) was held to be a D, while <1 was held to be an A”
Descriptive data on the participants (Table 1) was similar to that of the population of Yale pediatric residents at the time of the study. Interviews lasted an average of 9.8 minutes (SD: 4.2) and ranged from 5 to 20 minutes. For abstractors' identification of a content statement versus a noncontent statement, there were 8993 of a possible 9780 agreements (κ = .84). For “perfect agreement” in identifying an individual content code, there were 7003 of a possible 11415 (κ = .61). For “graduated agreement,” there were 7438 a possible 11415 (κ = .65).
Amount of Content
There were a total of 9780 statements over the 59 transcripts with 6038 coded as “content.” A total of 8416 content codes were applied during abstraction, including 1978 statements with 2 codes and 400 statements with 3 codes. The mean percent of statements devoted to content was 60.8% of the statements in each transcript (SD: 10.6%; range: 32.5%–80.0%). No differences were detected by the residents' gender, year in residency, or the disease being discussed.
All transcripts contained at least 1 statement on the following topics: the screening test, the infant now, the infant in the future, and background about the disease. At least 1 topic was missing in 26 transcripts (44.1% of the total). For example, 7 (11.9%) transcripts lacked statements about the parents or other family members now, and 10 (16.9%) lacked statements about the family in the future.
The most common communication topic was background about the disease (Table 2), although the information given to the residents implied that the infants were most likely heterozygous. Some transcripts included even more background: the greatest percentage was 56.8% in a transcript with 163 statements. No statistical difference was detected between the likely CF carrier and the SCH carrier interviews in the amounts of disease-background content. Within all of this background content, the 2 most common subtopics were “physiology, biochemistry, mechanism of action” (mean percentage of background content: 30.6% for the likely CF carrier and 42.1% for the SCH carrier scenarios), and “clinical features” (mean percentage of background content: 38.7% for likely CF carriers and 17.2% for SCH carriers). The average percentages of content devoted to other topics are shown in Table 2, although variability was seen here, too. For example, 11 transcripts included >15% of their content about parents and family in the future, and 3 transcripts included >15% in background about newborn screening. Another 3 transcripts included >15% of their content in background about genetics.
Quality Indicator 1: Key Content Items
The percentage of transcripts that contained each key content item, and the frequencies within each transcript are shown in Table 3. Not all of the key content items that were identified as “must-have” were equally common; the least commonly seen were the 3 items relating to genetic counseling for the future (discussion with relatives were missing in 88.1% of transcripts, future siblings are/may be at risk missing in 64.4%, and parental testing missing in 42.4%). The most common item among likely CF carrier transcripts was “infant needs/may need more tests” (97% of likely CF carrier transcripts), not unexpectedly given the variable implications of a positive CF newborn screening test. Fewer transcripts (64%) included a statement that the infant is (or may be) healthy or otherwise fine, although somewhat more (81%) promised that the infant will be (or will probably be) healthy or fine in the future.
The mean number of content items for likely CF carrier transcripts was 4.1 (SD: 1.2) of a maximum possible of 7 items. The mean number of content items for SCH carrier transcripts was 3.8 (SD: 0.9) of a maximum possible of 6 items. As shown in Fig 1A, only 5 (17%) likely CF carrier transcripts and 8 (27%) SCH carrier transcripts included the necessary number of content items for an A grade, whereas 9 (31%) likely CF carrier and 12 (40%) SCH carrier transcripts received the C or lower grade, qualifying for a hypothetical targeted intervention.
Other Content Items
Of the remaining content items, 4 of the 10 most common were from topic 9.0 (newborn screening background), including “newborn screening looks for many disorders” (68% of transcripts), “newborn screening uses a blood sample or heel prick” (66% of transcripts), “newborn screening looks at all newborns” (59%), and “newborn screening is done in the hospital or within days of birth” (54%). The other 6 most common were spread across the individualized topics, including “infant will be healthy” (56%), “infant has a single allele only” (47%), “mom or dad may be carriers” (59%), the nonspecific “a screening test positive” (53%), and the miscellaneous codes for “other things about infant's status in the future” (56%) and “other things about infant's adulthood reproductive issues” (51%).
A troubling finding was that 15 of 30 SCH carrier transcripts included statements about SCH carriers' risk for sudden death or severe illness. Many statements alluded to the military service data,32 but others referred to risk associated with severe exertion, high altitude, or “low oxygen tension.” Eleven transcripts contained 5 or more statements on the subject, and 5 included 15 or more statements. Qualitatively, it seemed that many such participants were attempting to soften the wording that they had used when they first brought up the issue. The ethical questions raised by this content item will be addressed further in Discussion.
Quality Indicator 2: Early Placement of Good News
Considering the likely CF carrier and SCH carrier transcripts together, the average point of appearance of the first good news content item was 25.2% of the way into the transcript (SD: 22.1%). Figure 1B shows the letter-grade definitions and data plus a histogram organized by deciles of time passed in the counseling episode (derived from the standardized percent-statement number). Sixteen (27%) transcripts had their first good news item appear in the first 10% of standardized statements, the condition necessary for an A grade. With the “C or less” threshold set at 30% or later, 17 (29%) transcripts were targets for a hypothetical intervention. One likely CF carrier transcript of these had no mention of possible carrier status or that the infant was healthy, the condition necessary for an F grade.
Quality Indicator 3: Background Content Ratio
Considering all of the transcripts together, the ratio was >1:1 (more background than individualized content) in 16 (27.1%) transcripts. The mean ratio of background to individualized content statements was 0.85:1 for likely CF carriers and 0.71:1 for SCH carriers. The difference between these 2 means was not statistically significant (1-way analysis of variance with a power at 81% for the observed difference).
Consistency of letter grades for the 3 quality indicators within each transcript yielded κ coefficients of .87 (1 vs 2), .70 (1 vs 3), and .70 (2 vs 3). The κ coefficients for consistency of participants across interviews were .79 for 1 and .67 for 2 but only .14 for 3 (see χ2 below).
A total of 6 individual transcripts received C or lower grades on 1 and 2 and an “excessive” score on 3. Grades for indicator 2 (early placement of good news) were significantly higher for SCH carrier cases than for the likely CF carrier cases (χ2, P < .05), whereas grades for quality indicator 1 (key content) were significantly higher for the likely CF carrier cases than for the SCH carrier cases (P < .05). Using the GPA system to compare indicator performance, the average GPA over the entire 59 transcripts was 3.0 (SD: 0.59), a B average. There was no statistically apparent difference in GPA on the basis of gender or year in residency.
If participants qualified for targeted feedback on the basis of C or lower grades on indicators 1 and 2 or “excessive” scores on indicator 3, then a total of 23 (79.3%) participants would have received feedback about at least 1 poor indicator performance. Of these participants, 1 had poor performance on all 6 possible indicators (ie, 3 indicators in both transcripts); 3 participants had poor performance on 4 indicators, 7 on 3, 3 on 2, and 9 on 1. Six (20.7%) participants had no poor indicator performances.
The purpose of this study was to develop and refine a set of tools to assess the “content” domain of communication quality after newborn screening for CF and SCH. Effective communication is important for genetic carrier screening because psychosocial complications have been traced to misunderstandings by family members or in the community.3 Although implications differ for the “SCH carrier” and the “likely CF carrier” screening results, we were able to tailor a single method to the communication demands of both situations (Tables 2 and 3 and Fig 1). The use of such adaptable methods in a quality assurance program after newborn screening may be able to improve communication processes and outcomes on a population scale.
In brief, this study found substantial variation in the content of participants' risk communication and in their scores on 3 communication quality indicators. Indicator performance was low enough to qualify for a C or lower grade in 35.6% of transcripts for indicator 1 and 30.5% for indicator 2, whereas 27.1% of transcripts were found to have “excessive” background on indicator 3. Another way to look at these results is to imagine the communication services delivered to 59 mothers of infants who were found via newborn screening to be carriers for SCH or CF. In this hypothetical group, 16 mothers would have heard too much background, 21 mothers would have heard less than half of the recommended key content items, and 18 mothers would not have heard the good news that their infant is a carrier or probably healthy until more than one third of the way into counseling. Although there is as yet no direct evidence between these measures of communication processes and outcomes, the lack of clear counseling provided to these mothers suggests that many of them could have been more likely to develop psychosocial complications or make misinformed reproductive decisions. This study has also provided an opportunity to demonstrate a new generation of communication assessment methods in action.
Several findings raised concern about whether some pediatricians understand what is important or appropriate for parents to hear. For example, the most common topic of communication was background about the disease, although the fictitious infants were unlikely to have a disease. This style of counseling, along with results from quality indicator 3, conflict with guidelines that information about the disease should be kept to a minimum for carrier counseling.31,33 Many transcripts lacked content about parents', siblings', or other relatives' risk or options for genetic testing. This finding may have been attributable partially to the lack of information about the mothers' and fathers' relatives, but it also suggests that some pediatricians and residents do not address a concept that is thought to be important by genetic counselors.14,17 Alternatively, the residents may have recognized that it was too early to raise this issue with parents, especially for the likely CF carriers who have a small chance of having 2 abnormal alleles.
In a more concerning example that also raises ethical questions about the balance of full disclosure against protecting the parent from potentially harmful or distracting information, we were alarmed at how many SCH carrier transcripts included content about sudden death or medical sequelae in SCH carriers. To us, this seems to be a topic of highly dubious value in the newborn period because of its potential for inducing parental anxiety or confusion about a very unlikely outcome. The alternative, however, is intentionally to withhold or delay counseling about an issue that could have considerable consequences for the child. Furthermore, many parents may run across a description of the sudden death “risk” from friends or on the Internet and come to believe that the clinician lied to them. Although no study has been made of this potential pitfall in counseling, our own opinion is that if clinicians choose to counsel parents of an SCH carrier infant about the infant's very small risk for symptoms, then they should be very careful to emphasize the low probability in advance of the information itself.32 Clinicians who choose to omit discussion of this issue during the newborn period should caution the parents that they are likely to hear odd things about sickle carrier status and ask them to bring in the information for discussion before becoming alarmed.
This study was limited by its small sample size, but some other specific limitations could be seen as strengths from a quality improvement perspective. For example, qualitative methods would have provided a richer description of the complex interchange between clinician and parent but would have had limited reliability and been prohibitively expensive for use on a population scale. Using standardized parent encounters instead of real parents avoids logistic, privacy, and consent difficulties associated with audiotaping actual clinic visits. Finally, concentrating on only a few clinician speech behaviors and our grounding in the simplistic “signal transmission” theory of communication34 may not be fully sufficient to achieve parental understanding but may have helped us to stay focused on clinician behaviors that are necessary for understanding.
One of our main goals for this study was to test the feasibility of a “communication quality assurance” method over a large population of clinicians. Population-scale communication methods are necessary now to optimize safety and outcomes for SCH and CF carriers. In the future, such methods may become even more important to work with screening tests that arise from the Human Genome Project or if other data prompt the American Academy of Pediatrics to restore support for routine carrier screening of children.4 Because quality improvement activities must have a lean budget and be able to be implemented consistently by many different project personnel, we tracked time and expenses for the BSCA interviews and abstractors. This helped us to estimate a budget for future projects of <$50 per participant (most of which would be transcription) and <5 minutes of abstractor time per indicator. Reliability for use of the data dictionary in this study was good; the coefficients between .61 and .87 are respectable for social science methods at an early stage of development. Given our inter-transcript reliability, there would be a reasonable chance that without intervention, these participants would score similarly on future transcripts. Our approach will also help to target interventions to those with the most to gain. For example, using the C or lower threshold in this project would have resulted in 23 (76.7%) of 30 pediatric residents' receiving the hypothetical targeted intervention regarding at least 1 of their indicators. In contrast to such a targeted approach, an educational intervention would have cost much more because it would typically be provided to all students. Feedback from explicit measures can also be more behavior specific than that derived from the structured implicit scales that currently are in use. Thus, our method should provide a low-cost mechanism to rank quantitatively and reliably clinicians' performance at communication tasks, allowing comparison, targeting, feedback, and benchmarking for improvement.
In summary, this study suggests that communication quality may be characterized by moments of excellence surrounded by missed opportunities and an unfortunate emphasis on background. Future projects will investigate “conduct” indicators for this same data set and will replicate findings in larger, community-based data sets for other newborn screening tests. Communication about newborn screening is challenging because of differences in testing methods, diseases, and families,22 and psychosocial complications of poor communication in part are inhibiting society from implementing new screening programs. This study offers part of a low-cost, quantitatively reliable method to measure processes of communication after newborn screening in the real world. If integrated into US newborn genetic screening practices, then such a system may benefit clinicians and families and help satisfy society's ethical concerns about the risk and benefit with expanded newborn screening and genetic screening.
Dr Farrell is supported in part by grant 7K01HL072530 from the National Heart, Lung, and Blood Institute.
An earlier, partial version of this analysis first was presented as a student thesis by Ms La Pean.
When the project was begun, Dr Farrell was assistant professor and Ms Ladouceur was research associate II at the Yale Primary Care Medicine Research Center (Waterbury, CT); Ms La Pean was a master's student in the genetic counseling program at Northwestern University (Chicago, IL).
We are grateful to Jeffrey Stein, MD, and Laura Farrell, MAT, for invaluable advice and assistance.
- Accepted April 8, 2005.
- Reprint requests to (M.H.F.) Internal Medicine & Pediatrics, Center for Patient Care and Outcomes Research, 8701 Watertown Plank Rd, PO Box 26509, Milwaukee, WI 53226-0509. E-mail:
No conflict of interest declared.
- ↵National Newborn Screening and Genetics Resource Center. National Newborn Screening Report—2000. Austin, TX: National Newborn Screening and Genetics Resource Center; 2003. Available at: genes-r-us.uthscsa.edu/resources/newborn/00chapters.html
- ↵Markel H. Scientific advances and social risks: historical perspectives of genetic screening programs for sickle cell disease, Tay-Sachs disease, neural tube defects, and Down syndrome, 1970–1997. In: Holtzman NA, Watson MS, eds. Promoting Safe and Effective Genetic Testing in the United States: Final Report of the Task Force on Genetic Testing. Baltimore, MD: The Johns Hopkins University Press; 1998:161– 176
- ↵Nelson RM, Botkjin JR, Kodish ED, et al. Ethical issues with genetic testing in pediatrics. Pediatrics.2001;107 :1451– 1455
- Kessler S. Psychological aspects of genetic counseling. XIII. Empathy and decency. J Genet Couns.1999;8 :333– 344
- ↵Marteau TM, Croyle RT. The new genetics. Psychological responses to genetic testing. BMJ.1998;316 :693– 696
- ↵Ciske DJ, Haavisto A, Laxova A, Rock LZ, Farrell PM. Genetic counseling and neonatal screening for cystic fibrosis: an assessment of the communication process. Pediatrics.2001;107 :699– 705
- ↵Andrews LB, Fullarton JE, Holtzman NA, et al, eds. Assessing Genetic Risks: Implications for Health and Social Policy. Washington, DC: National Academy Press; 1994
- Baile WF, Buckman R, Lenzi R, Glober G, Beale EA, Kudelka AP. SPIKES—A six-step protocol for delivering bad news: application to the patient with cancer. Oncologist.2000;5 :302– 311
- ↵Baker Diane L, Schuette Jane L, Uhlmann Wendy R. A Guide to Genetic Counseling. New York, NY: Wiley-Liss; 1998
- Committee on Risk Perception and Communication, Commission on Behavioral and Social Sciences and Education, and Commission on Physical Sciences, Mathematics, and Resources, United States National Research Council. Improving Risk Communication. Washington, DC: National Academy Press; 1989
- Kurtz S, Silverman J, Draper J. Teaching and Learning Communication Skills in Medicine. Abingdon, Oxon, UK: Radcliffe Medical Press; 1998
- ↵Gregg RG, Simantel A, Farrell PM, et al. Newborn screening for cystic fibrosis in Wisconsin: comparison of biochemical and molecular methods. Pediatrics.1997;99 :819– 824
- ↵Hoppe RB. Standardized (Simulated) Patients and the Medical Interview. In: Lipkin M, Putnam SM, Lazare A, eds. The Medical Interview: Clinical Care, Education, and Research. New York, NY: Springer; 1995:397– 404
- ↵Neuendorf KA. The Content Analysis Guidebook. Thousand Oaks, CA: Sage; 2002
- ↵Glaser BG, Strauss AL. The Discovery of Grounded Theory. Chicago, IL: Aldine; 1967
- ↵Feinstein AR. Clinical Epidemiology: The Architecture of Clinical Research. Philadelphia, PA: WB Saunders; 1985
- ↵Falvo DR. Effective Patient Education: A Guide to Increased Compliance. 2nd ed. Gaithersburg, MD: Aspen; 1994
- ↵La Pean A, Farrell MH. Initially misleading communication of carrier results after newborn genetic screening. Pediatrics.2005;116:1499–1505
- ↵Parsons EP, Clarke AJ, Bradley DM. Implications of carrier identification in newborn screening for cystic fibrosis. Arch Dis Child Fetal Neonatal Ed.2003;88 :F467– F471
- ↵Shannon CE, Weaver W. Mathematical Theory of Communication. Urbana, IL: University of Illinois Press; 1949
- Copyright © 2005 by the American Academy of Pediatrics