Original Contribution

Pediatric health screening and referral in the ED

Leslie S. Zun MD, MBA^a,b,*, LaVonne Downey PhD^c

^aDepartment of Emergency Medicine, Rosalind Franklin University of Medicine and Science, Chicago Medical School,

IL 60069, USA

^bDepartment of Emergency Medicine, Mount Sinai Hospital, IL 60608, USA

^cRoosevelt University, Chicago, IL 60605, USA

Received 17 November 2004; revised 8 January 2005; accepted 25 February 2005

Abstract

Introduction: Many studies have demonstrated the importance of performing preventive care in the ED. The primary objective of this study was to identify unmet health needs in the ED of the pediatric patient population. The secondary objective was to determine if the patient’s parent or guardian would accept health referrals and bring the child to follow up with a doctor.

Methods: Age- and sex-specific algorithms concerning preventive care were developed from the US Department of Public Health Clinicians’ Book of Preventative Health. A convenience sample of patients and their families who presented to the ED were asked to participate in the study. The exclusion criteria consisted of patients who were institutionalized, unstable, and had parents who were unable to communicate or declined to participate. After 1 week, the parents were followed up by telephone to find out if they had made an appointment with a doctor as recommended. One month after the ED visit, the health care’s computer system was queried to confirm that the appointment had been completed. Data were analyzed using SPSS (version 10.0; SPSS, Chicago, Ill), and tests of significance used were the Pearson v², frequency test, and crosstabs. This study was institutional review board-approved as exempt. Results: Two hundred three pediatric patients were enrolled. Most of the patients had a primary care physician (87.1%, 176/203) and insurance (85.6%, 172/203). Only 25 (12.3%) of 203 needed any referrals, with an acceptance rate of 72.0% (18/25) and completion rate of 40% (10/25). The most frequent unmet need was for urine, lead, and anemia screening (19.4%, all 20/103). Few patients needed immunizations (1.0%, 2/203), alcohol screening (2.0%, 2/100), or blood pressure testing (3.0%, 3/100). Analysis of the correlation between getting 1 or more referrals and race was found to have a significant relationship (v = 19.69, df = 6, P = .003) but not with sex, age group, insurance, or primary care physician ( P N .05).

Conclusion: In this study, 12.3% of the patients were found to have unmet health care needs. Insurance status had no bearing on the need for referrals. Assessment in the pediatric population for unmet health care needs was found to be a low-yielding, labor-intensive process.

D 2005

Presented at the Second Annual Mediterranean Emergency Medicine Conference in Barcelona, Spain, in 2003.

T Corresponding author. Department of Emergency Medicine, Mount Sinai Hospital, 15th and California, Chicago, IL 60608, USA.

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2005.02.049

Introduction

According to Healthy People 2010, section 14.24a, an important goal is to increase the proportion of young children who receive vaccines [1]. Immunization rates vary from state to state, from urban localities vs rural ones, and by ethnic groups. Various authors have found that the immunization rate of children is significantly less than the mandated 100%. In 1998, 73% of children received all vaccines recommended for universal admin- istration [2]. National statistics for immunizations among children aged 19 to 35 months varies in urban areas from 63% to 87% [3]. McConnochie and Roghmann [4] found that the 27% of the youths whom they evaluated were missing 1 or more immunizations at age 4.4 years. They also found that ED records accounted for 18% of missed immunization opportunities.

Numerous authors have proposed preventive screening, testing, and treatment in the ED. The Society of Academic Emergency Medicine’s Public Health and Education Task Force generated a list of targeted ED interventions aimed at patients in the general and high-risk groups: screening, counseling, immunization, chemoprophylaxis, health pro- motion material, social services, and ED surveillance [5]. The Society of Academic Emergency Medicine’s Public Health and Education Task Force on Preventative services found from evidence-based review that alcohol screening and intervention, HIV screening, identification of hyperten- sion, pneumonococcal vaccination, smoking cessation, and social services needs assessment and referral were useful [6]. Williams and others [7] found that, although emergency physicians feel responsible for promoting the health of their patients, few routinely screen and counsel patients on prevention and many were not confident of their ability to help their patients with respect to preventive care. Contro- versy exists as to the role of the ED in pediatric immunizations. Szilagyi and others [8] found that 30% of the surveyed pediatricians and family physicians recom- mended that immunizations be administered in the ED.

In response to the controversy, we postulated that identification and referral would be preferred to immuniza- tion and treatment in the ED. The primary purpose of this study was to determine if pediatric patients have unmet health care needs, as determined by the administration of a health screening and referral tool. The secondary purpose was to determine if the child would follow up with the clinic or doctor referrals.

Methods

The guidelines of the US Public Health Service were followed for age- and sex-specific screening and referral needs to develop algorithms for patient referrals [9]. The algorithms were age- and sex-based and questioned the patients’ parent or guardian about key examinations or

immunizations. Patients were assigned to 1 of 3 groups based on age and sex: 0- to 10-year-old males/females, 11- to 18-year-old males, and 11- to 18-year-old females. Age and sex were the basis for whether to inquire about blood pressure, vision and hearing screening, immuniza- tions, anemia, lead screening, sexually transmitted disease exposure, alcohol and substance use, tuberculo- sis, and cigarette smoking (Appendix A). The CAGE assessment tool for alcohol and drug use was used as the means to determine substance abuse problems [10,11]. The patients were referred to 1 of 3 resources: to their private physician, to a multispecialty group practice for unassigned patients with insurance coverage, or to a Family Medicine clinic for unassigned patients without coverage. The patient or patient’s care provider (ie, parent or guardian) was used as a source material with respect to the reporting of health care compliance.

A convenience sample of patients was approached, primarily during daytime, Monday through Friday, to determine their willingness to being interviewed for a health screening and Referral program. The study was conducted in a level 1 pediatric and adult trauma center with a total of 45000 annual ED visits with approximately 25% pediatric patients. The ED is located in an Inner city, African American and Hispanic neighborhood, with 40% of the patients on public assistance, 40% of the patients without any insurance, and 10% with commercial or managed care coverage. The inclusion criteria were all stable pediatric patients not in need of immediate intervention. The exclusion criteria eliminated institutionalized or unstable patients and those parents or guardians who were unable to communicate or refused to participate.

Research fellows administered the survey tool to the parents or guardians who agreed to participate in the health- screening program. Patients and their guardians were told that involvement in this program was voluntary and at any time they could stop or withdraw from the study. Partic- ipants were told bWelcome to the Emergency Department. This survey is a health screening and referral interview. It will take about 10 to 15 minutes to complete. You will be asked questions about your physical health and the health- care that you receive. After you have answered all of the questions, you will get your results and referral recommen- dations. If you agree, we will assist you in obtaining referrals and a copy of this survey will be given to you. Involvement in this program is voluntary and at any time you can stop or withdraw from the study.Q The research fellow completed the patient data collection sheet and selected the correct algorithm for the patients’ sex and age group. If the patient had a doctor, they get referred back to their doctor with the referral sheet. If the patient had any sort of insurance (including public aid), they will get a referral to a system clinic, and if they did have any insurance coverage, the patient will be referred to a federally qualified health care clinic. The patient is given a copy of the results, with the referrals listed for each medical problem. The patient is

called after 1 week to determine whether the patient has made an appointment. One month later, the health care system computer was queried to determine if the patient followed up with a health care provider in the health care system. The health care system computer provides visit history for patients seen in the hospital, outpatient clinics, and 50 Primary care clinics associated with the health care system, including those that are federally qualified.

Table 2 Referrals

Values are expressed as % (n).

Overall		Aged 0-10 y	Aged 11-17 y
No. of referrals 0	87.6 (178/203)	87.4 (90/103)	88.0 (88/100)
1-3	11.9 (24/203)	11.7 (12/103)	12.0 (12/100)
4-8	0.5 (1/203)	9.7 (1/103)	0.0 (0/100)
AcceptED referral Yes 72.0 (18/25)		92.3 (12/13)	55.5 (6/11)
No	20.0 (5/25)	0.0 (0/13)	45.5 (5/11)
Unknown	8.0 (2/25)	7.7 (1/13)	0.0 (0/11)

The data were inputted into an SPSS file (SPSS, version 10.0; Chicago, Ill) based on the number of items that

screened positive, the primary care provider, demographic data, need for services, agreement to the assessment and	Made appointment Yes 40.0 (10/25)		61.5 (8/13)	16.7 (2/12)
referral, types and number of referrals, acceptance of	No	44.0 (11/25)	30.8 (4/13)	58.3 (7/12)
referrals, and appointments made with physicians. A convenience sample was obtained to have a sample of at	Uncertain 16.0 (4/25) Confirmed appointment		7.7 (1/13)	25.0 (3/12)

least 50 patients in each age/sex group. The data were analyzed to determine if there were any correlations between age, sex, insurance coverage, health care needs and referrals, number and acceptance of referrals, and actual follow-up. Age and sex groups were lumped in the analysis because of the small number of patients with unmet health care needs. Pearson v² analysis was used because of the categorical nature of most variables. The study was institutional review board-approved as exempt because it is for an bimprovement in the current provision of medical care.Q

Results

A total of 210 patients in the ED were approached for the study. The parents or guardians of 7 patients refused to enroll in the study, 3 stated they were not interested, and 4

Overall	Aged 0-10 y	Aged 11-17 y
(N = 203)	(n = 103)	(n = 100)

Confirmed	40.0 (10/25)	53.9 (7/13)	25.0 (3/12)
No follow-up	48.0 (12/25)	23.1 (3/13)	75.0 (9/12)
Unable to find	8.0 (2/25)	23.1 (3/13)	0.0 (0/12)

refused to sign Health Insurance portability and accounta- bilty consent form. The demographic profile of the patients surveyed consisted of the following: half were males; 89 (46.1%) of 203, African American; 101 (51.8%) of 203,

Hispanic; and 2 (1.0%) of 203, white (Table 1). Most of the patients had a primary care physician (87.1%, 176/203) and insurance (85.6%, 176/203).Only 25 (12.3%) of 203 needed any referrals, with an acceptance rate of 72.5% (18/ 25) and completion rate of 40.0% (10/25) (Table 2). The most frequent unmet need was for urine, lead, and anemia screening (19.4%, all 20/103) (Table 3). Few patients needed immunizations (1.0%, 2/203), alcohol screening (2.0%, 2/100), or blood pressure testing (3.0%, 3/100).

Because of the categorical and dichotomous nature of most variables, a Pearson v² test was used. Analysis of the correlation between getting 1 or more referrals and race revealed a significant relationship (v = 19.69, df = 6, P =

Table 1 Demographic results

Sex Male	101 (50.0)	51 (49.5)	50 (50.0)	.003) but not with sex, age group, insurance status, or
Female Missing data	101 (50.0) 1	52 (50.5)	50 (50.0)
Race AA	89 (46.1)	37 (35.9)	52 (52.0)	Urine screening	19.4 (20/103)
Hispanic	101 (51.8)	61 (59.2)	40 (40.0)	Lead	19.4 (20/103)
White	2 (1.0)	1 (9.7)	1 (1.0)	Anemia testing	19.4 (20/103)
Other	2 (1.0)	2 (1.9)	0 (0.0)	Hearing testing	14.6 (15/103)
Missing	9 (4.5)	2 (1.9)	7 (7.0)	Vision testing	14.6 (15/103)
PCP				Papanicolaou smear	10.0 (2/20)
Yes	176 (87.1)	95 (92.2)	81 (81.0)	Sexually transmitted disease	10.0 (3/29)
No	27 (12.9)	8 (7.8)	19 (19.0)	Cigarettes	10.0 (10/100)
Insurance status				Drug Abuse	4.0 (4/100)
Yes	172 (85.6)	96 (93.2)	76 (76.0)	Blood pressure	3.0 (3/100)
No	23 (10.9)	7 (6.8)	16 (16.0)	Tuberculosis	3.0 (6/203)
Unknown	7 (3.5)		8 (8.0)	Alcohol abuse	2.0 (2/100)
Values are expressed as n (%). AA indicates African American; PCP, primary care physician.				Immunizations (aggregated) 1.0 (2/203) Values are expressed as % (n).

Table 3 Rank order of positive screenings

Referral, yes or no, and
Sex	4.115	4	.391
Race 19.69 6			.003
Age group	19.839	18	.342
Insurance	7.966	6	.241
PCP	5.175	3	.159
Acceptance of referrals and
Sex	0.190	2	.909
Race	19.442	4	.054
Age group	17.341	12	.137
Insurance	10.606	4	.031
PCP	2.831	2	.243
Followed up and
Sex	2.114	3	.549
Race	15.412	8	.052
Age group	21.084	24	.634
Insurance	12.925	8	.114
PCP	15.412	8	.052

primary care physician ( P N .05) (Table 4). Acceptance of the referral was correlated with insurance status (v = 10.606, df = 4, P = .031) and not with sex, age group, race, or primary care physician ( P N .05). None of the variables were correlated with actual follow-up ( P N .05).

Table 4 Correlations

v df P

Discussion

Contrary to the findings of several other studies, this study demonstrated that only a relatively small population of ED pediatric patients has unmet health care needs. We cannot explain why this study had a low rate of needed immunizations when other authors found that the unimmu- nized rate was 27% [4]. The study found relative need for urine, anemia, and lead screening (19.4%). Therefore, screening of ED pediatric patients overall in this study had a low yield.

The study was performed by research fellows rather than ED staff because of their limited time in a busy level 1 trauma center. We found that continuation of the screening and Referral process, either by the ED staff or research personnel, would be labor-intensive and costly. There may be value in determining the health care needs of a population using less costly means. A recent study at the University of Chicago demonstrated that ED patients are willing and interested in participating in a self-assessment and health education study while waiting to be seen in the ED. The major goal of this program was to demonstrate that patients would be willing to be educated on health risk topics [12]. Another study used a computer-directed assessment of patients’ medical needs before surgery [13]. Kempner [14] used a self-administered questionnaire for psychosocial screening in the pediatric age population. Lutner et al [15] found that patients would use a small

handheld device to answer health-related questions. It is uncertain whether the conversion of the staff-administered health screening and referral program to a self-directed computer program would be used and would alleviate the cost involved in administering the program. Gregor and others [16] used an interactive computer program in the ED in an attempt to prevent alcohol misuse among adolescents. The study was based on the recommendations promul- gated by the US Department of Public Health’s Task Force on Preventive Care [9]. However, there is some dispute as to the proper preventive recommendations for children. Elstel [17] performed an analysis of 5 organizations and found that all groups recommended immunizations and screening for health issues such as hypertension, obesity, and Tobacco use. We did not survey for birth-related testing for hemoglobinopathy, phenylalanine level, thyroxine resin uptake, thyroid-stimulating hormone, or interventions for high-risk populations. The study did not address the need for counseling for injury prevention, sexual behavior, diet and exercise, substance use, or dental care. We did not screen for HIV/AIDS and Violence exposure or victimiza- tion because they were not among the recommendations of the US Department of Public Health, despite the epidemic levels of those diseases in the community studied. Although the National Depressive and Manic-Depressive Association recommend age-appropriate screening and diagnosis of children and adolescents with these disorders, they were not addressed in this study [18]. It would be valuable to develop Screening tools and determine the best means to assess for these conditions that are prevalent in the

community served by the sample ED population.

The health care assessments were based on the US Department of Public Health recommendations [9]. Not all the recommendations were sufficiently straightforward; several required some interpretation. Standard questions to perform the health care assessment were not found in the recommendation and were developed for the survey. The survey tool used in this study was not validated. The assessment for alcohol and substance abuse was not provided in the recommendations; therefore, the CAGE assessment was used.

There are many other limitations to this study. The convenience sample of the patients in the inner city may not properly represent the general population in the United States. The study was performed primarily during the weekdays, limiting its generalization to the rest of the week. The study was limited by the number of respondents in each group for analysis purposes. Parents and guardians who did not respond or refused to enroll in the study also limited the value of the results. The study was dependent upon the parents or guardians to provide the source information concerning immunizations and other health care issues. However, Goldstein and others [19] found that there were a large number of inaccurate assessments of children immu- nization status in the ED. In a study with a similar Conceptual framework, Vaughan and others [20] screened

youth in schools and referred the youth to a clinic for follow-up. The tool that they developed had a moderate amount of false positives through single-item identification. Referrals were made to services in the health care system, but the quality of the services provided was not determined. The study did determine if the appointment was made but did not determine if the appointment was completed and whether the required test or immunization was actually performed. The study used an unvalidated set of questions to assess the patients’ unmet health care needs.

Conclusion

Few patients in this study proved to have had unmet health care needs. This study demonstrated that screening pediatric patients in the ED for unmet health care needs is resource-intensive and has a low yield, providing an unfavorable Cost benefit ratio. The study was based on the interpretation of US Public Health Service criteria for age- based health care needs.

Acknowledgments

We thank Olga Borisovsky for her assistance with patient enrollment.

Appendix A. Screening performed by age and sex

References

1. Healthy people http://www.healthypeople.gov. Accessed 9/6/03.
2. CDC. National, state, and urban area vaccination coverage levels among children aged 19-35 months. http://www.cdc.gov/mmwr/ preview/mmwrhtml/mm492a1.htm [Access 9/6/03].
3. Gest KL, Magrgolis P, Bordley C, Stuart J. Measuring the process of preventive service delivery in primary care practices for children. Pediatrics 2000;106:879 - 85.
4. McConnochie KM, Roghmann KJ. Immunization opportunities missed among urban poor children. Pediatrics 1992;89:1019 - 26.
5. Rhodes KV, Gordon JA, Lowe RA. Preventative care in the emergency department, part I: clinical preventative services-are they relevant to emergency medicine? Acad Emerg Med 2000;7: 1036 - 41.
6. Irvin CB, Wyer PC, Gerson LW. Preventive care in the emergency department, part II: clinical Preventive services-an emergency medicine evidence-based review. Acad Emerg Med 2000;7:1042 - 54.
7. Williams JM, Chinnis AC, Gutman D. Health promotion practices of emergency physicians. Am J Emerg Med 2000;18:17 - 21.
8. Szilagyi PG, Rodewald LE, Humiston SG, et al. Immunization practices of pediatricians and family physicians in the United States. Pediatrics 1994;94:517 - 23.
9. US Public Health Service. The clinician’s handbook of preventative services. 2nd ed. Germantown, (Md)7 International Publishing; 1998.
10. Bush B, Shaw S, Cleary P, Delbanco TL, Aronson MD. Screening for alcohol abuse using the CAGE questionnaire. Am J Med 1987;82: 231 - 5.
11. Ewing JA. Detecting alcoholism: the CAGE questionnaire. JAMA 1984;252:1905 - 7.
12. Rhodes K, Lauderdale D, Stocking C, Howes D, Roizen M, Levinson W. Better health while you wait: a controlled trial of a computer-based intervention for screening and health promotion in the emergency department. Ann Emerg Med 2001;37(3):284 - 91.
13. Raines JR, Ellis LBM. A conversational microcomputer based health risk appraisal. Comput Programs Biomed 1987;14:175 - 84.
14. Kempner KJ. Self-administered questionnaire for structured psycho- social screening in pediatrics. Pediatrics 1992;89:433 - 6.

Aged 0-10 y, male/female

Aged 11-17 y,

male

Aged 11-17 y,

female

1. Lutner R, Roizen M, Stocking C, Thisted R, Kim S, Duke P, et al. The automated interview versus the personal interview. Anesthesiology 1991;75:394 - 400.
2. Gregor MA, Shope JT, Blow FC, et al. Feasibility of using an

Tuberculosis X X X Vision testing X

Hearing testing X Anemia testing

interactive laptop program in the emergency department to prevent alcohol misuse among adolescents. Ann Emerg Med 2003;42:276 - 84.

1. Elstel AB. Comparison of recommendations for adolescent clinical preventive services developed by national organizations. Arch Pediatr Adolesc Med 1998;152:193 - 8.

Sexually transmitted

X [18] Lewis L. Unmet needs in diagnosis and treatment of Mood disorders in children and adolescents. Biol Psychiatry 2001;49:960 - 1.

1. Goldstein KP, Kviv FJ, Daum RS. Accuracy of immunization histories provided by adults accompanying preschool children to a pediatric emergency department. JAMA 1993;270:2190 - 4.

disease
Cigarettes		X	X
Blood pressure		X	X
Alcohol abuse		X	X
Drug abuse		X	X
DPT	X	X	X
Polio	X
MMR	X
HiB	X
Hepatitis	X
Varicella	X

1. Vaughan RD, Mccarthy JF, Walter HJ. The development, reliability and validity of a risk factor screening survey for the urban minority junior high school students. J Adolesc Health 1996;19:171 - 8.

DPT indicates diphtheria-pertussis-tetanus; MMR, mea- sles, mumps, and rubella; HiB, Haemophilus influenzae type B; X, used in screening test.