Children With Hearing Loss and Increased Risk of Injury

There is little research on the health and healthcare of people with birth or childhood onset hearing loss. Adults who became deaf before age 3yrs (called “prelingual”) have poorer health and less frequent clinician visits than non-deaf adults [1]. Qualitative research [2] and expert opinion [3] report some prelingual deaf adults use emergency department services frequently or inappropriately. It is possible that the healthcare behaviors of prelingual deaf adults have roots in their childhood health and healthcare experiences. I am not aware of health research with non-deaf adults who have birth- or childhood-onset unilateral or less severe bilateral hearing loss.

Mann et al [4] give a glimpse of the health and healthcare behaviors of some families with a child with hearing loss. In their secondary analyses of claims data, they found that a group of Medicaid enrollees with at least two episodes of a hearing loss diagnosis code had more emergency department visits and more treated injuries than a comparison group. I agree with the authors that these findings warrant more research. The authors discuss the limitations of their research in the article. I comment on two of these limitations here (use of billing records and potential confounders) because they can help shape future research. My two research questions are: 1) For children who are deaf or hard of hearing, how do children with a diagnosis code for hearing loss compare with those without the diagnosis code? 2) Are the factors related to clinicians’ decision to use a hearing loss diagnosis code also associated with the likelihood of injury treatment?

The authors mention the challenges of using billing data to identify children with hearing loss. I suspect that most Medicaid children with hearing loss were not identified in the data. Before exclusions, Mann et al found 4,647 South Carolina Medicaid children with a hearing loss diagnosis code in 2003. South Carolina Medicaid insures 300,000 [5] to 500,000 [6] children, so the prevalence of a hearing loss diagnosis code is at most 1.5%. This is far below the 11% prevalence of any degree of childhood hearing loss cited in the article’s Introduction (the prevalence of hearing loss is probably higher in the Medicaid population, since, as the authors point out, childhood hearing loss is associated with low income). Since most Medicaid children with hearing loss would not have a diagnosis code for hearing loss, it would be helpful to know which children with hearing loss get a hearing loss diagnosis code.

For respondents to the Medicare Current Beneficiary Survey, hearing loss diagnosis codes are not a good way to identify people with hearing loss: 81% of respondents who reported a hearing loss did not have a claim with a hearing loss diagnosis code, and 63% of respondents with a hearing loss diagnosis code reported no trouble hearing [7, Table 3]. I have not seen similar data for children or Medicaid. With future research we need to study clinicians’ decisions about the hearing loss diagnosis codes before we can reliably use these codes to answer questions about people with hearing loss.

Knowing about clinicians’ decisions regarding the use of hearing loss diagnosis codes would affect how I interpret the study’s findings. For example, although hearing loss diagnosis codes do not indicate severity of hearing loss, it is possible that clinicians are more likely to use a hearing loss diagnosis code with a child who is deaf or has profound bilateral hearing loss. This would be consistent with the authors’ speculation about the mechanism of risk for injury treatment in children with a hearing loss diagnosis code (limited detection of environmental sounds and sub-optimal communication with parents). However, it is not consistent with most children with hearing trouble: 38% have unilateral hearing trouble and 32% have bilateral hearing trouble but “can hear words spoken or whispered” [8, Table 1]. Knowing if injury risk was highest in children with more severe hearing loss would help me to focus prevention efforts in the community and in clinical practice, without alarming families with a child who may not be at increased risk for injury.

Clinician decisions to use a hearing loss diagnosis code may be related to factors that also are related to injury risk. For example, clinicians who see more patients with hearing loss may be more likely to use a hearing loss diagnosis code. Hearing loss is more prevalent outside of a Metropolitan Statistical Area (MSA) [8, Table 9], and rural children are at increased risk for injury [9]. If use of a hearing loss diagnosis code is related to rural location, that might explain some of the increase in injury treatment associated with hearing loss diagnosis codes in children. Future research could adjust analyses for the hospital’s or the child’s ZIP code or county to see if some of the difference between the study and comparison group may be related to rural location.

Another possibility is that clinicians may be more likely to use hearing loss diagnosis codes with patients with severe comorbidities, such as those who require institutional level care. The authors report that the study group is 15 times more likely than the comparison group to qualify for Medicaid because of the need for institutional level care. Hearing loss alone would not require institutional level care, so it seems that the reason for qualifying for Medicaid (income vs. institutional level care) could function as a measure of comorbidity. The authors excluded children with some known comorbidities from their analyses. I am curious how analyses adjusted for institutional level care would change the relative risk for injury treatment and use of emergency department services for children with a hearing loss diagnosis code.

I agree with the authors and the comments from Drs. Osterling and Xiang that we need further research on the health of children with hearing loss. If that research will continue to use billing data, then we need to study clinician decision-making as it relates to hearing loss diagnosis codes. Consistent with Dr. Xiang’s comments, future research would be most useful for family physicians if it examined health and healthcare in the context of families. For example, perhaps having a child who is deaf or hard of hearing increases stress in the family such that all the children in that family, regardless of hearing status, are at increased risk for injuries or poor health outcomes. Alternatively, perhaps having a child who is deaf or hard of hearing helps to focus a family on health and healthcare such that all children in that family have frequent clinician visits and perhaps better adherence with preventive services such as immunizations. Family level research also should examine parents’ health and healthcare; family caregiver’s health is an emerging public health issue [10]. Working with those research results, clinicians, deaf individuals, their families, public health officials, and other interested parties could design interventions to promote health and prevent illness or injury in families with a member who is deaf or hard of hearing.

References

1. Barnett S, Franks P. Health care utilization and adults who are deaf: Relationship with age at onset of deafness" Health Services Research. 2002;37 (1), 103–118.

2. Steinberg A, Barnett S, Meador H, Wiggins E, Zazove P. (2006). Healthcare system accessibility: Experiences and perceptions of Deaf people. Journal of General Internal Medicine. 2006;21(3):260-266.

3. Rochester Deaf Health Task Force Report. 2004. Finger Lakes Health Systems Agency, Rochester NY. http://www.flhsa.org/pubs/Deaf%20Health%20Task%20Force%20Report.pdf (Accessed December 2, 2007).

4. Mann JR, Zhou L, McKee M, McDermott S. Children with hearing loss and increased risk of injury. Ann Fam Med. 2007;5(6):528–533.

5. South Carolina - Kaiser State Health Facts. http://www.statehealthfacts.org/mfs.jsp?rgn=42&rgn=1 (Accessed December 2, 2007).

6. 2005 South Carolina Kids Count Report. http://www.sckidscount.org/kc05.asp?COUNTYID=47 (Accessed December 2, 2007).

7. Identification of Performance Standards for The Deaf and Hard of Hearing: Report on the Analysis of Data Sources for Assessing the Health Status of Deaf and Hard of Hearing People. A Report to The Centers for Medicare & Medicaid Services, prepared by Delmarva Foundation for Medical Care, Inc, Easton, MD and Gallaudet Graduate Research Institute, Gallaudet University, Washington, DC. March 7, 2002.

8. Ries, Peter W. Prevalence and characteristics of persons with hearing trouble: United States, 1990-91. National Center for Health Statistics. Vital Health Stat 10(188). 1994. http://cdc.gov/nchs/data/series/sr_10/sr10_188.pdf

9. Safe Kids Worldwide (SKW). Rural Safety. Washington (DC): SKW, 2007. http://www.usa.safekids.org/content_documents/2007_Fact_Sheet_Rural_Safety.doc (accessed December 2, 2007).

10. Talley RC, Crews JE. Framing the public health of caregiving. Am J Public Health. 2007;97:224–228.

Competing interests:   None declared

We would like to thank Dr. Xiang for his comments, and for summarizing his own work in the are of injury epidemiology in children with disabilities. We agree that an important strength of our study was our ability to control for general propensity to seek treatment, as measured by the number of visits for non-injury reasons.

We also acknowledge Dr. Xiang's point about not being able to examine the psychosocial context of the children with and without hearing loss, and how non-medical factors may have impacted injury treatment rates. We agree that it is important to look beyond the simple medical "diagnosis" of a disability and to examine family, environmental, and other factors that may be important. Unfortunately, since we utilized Medicaid billing data for this study, we were limited in the kinds of factors we could assess. Additional research on the psychosocial factors underlying injury risk in children with disabilities is indeed needed.

Competing interests:   None declared

I read with great interest the article by Mann et al in the November/December issue of the Annals of Family Medicine reporting the increased injury risk facing children with hearing loss (1). Using South Carolina Medicaid billing data for 2002-2003, the authors compared rates of emergency department or hospital treatment for injury among children with and without a diagnosis of hearing loss and concluded that children with hearing loss face an increased risk of injury when compared to children without hearing loss.

Previous research on injury among children with disabilities has produced similar results (2-5). My own research team analyzed data from the 2000-2002 National Health Interview Survey and found that children 5- 17 years of age with a vision/hearing disability had a significantly higher rate of medically treated injuries than did healthy children in the same age range (4.2% vs. 2.5%). In addition, the adjusted odds ratio was similar to the adjusted relative rate reported by Mann et al (3). In another study (to be published in an upcoming article), we found the injury prevalence ratio to be 1.78 (95% confidence interval, 0.95-1.75) among children with hearing problems compared to children without disabilities (4). A similar pattern of increased incidence was found in an analysis of Ohio Medicaid billing data for burn injuries among Ohio Medicaid children with disabilities (5). Scientific findings from research efforts by previous researchers and Mann et al indicate that safety issues and injury prevention are an important public health problem facing millions of children with disabilities in the United States.

I want to praise Mann et al for using rigorous methods to analyze the rather complex Medicaid billing data. In particular, I would like to commend them for testing whether the elevated rates of injury among children with hearing loss could be due to a potential tendency of that population to seek more emergency or hospital treatment. After controlling for non- injury-related emergency department visits and hospitalizations, Mann et al found that children with hearing loss still had higher rates of injury treatment.

Another potential problem of injury studies that rely on self- reported interview data is reporting differences between groups. Among children with disabilities this means that the injury-reporting behaviors of parents of children with disabilities could be significantly different than those of parents of children without disabilities. To the best of my knowledge, no study has rigorously examined this important issue. More research is needed to understand the impact of medical care seeking behaviors and recall bias on the conclusions of injury research among children with disabilities.

While I generally agree with earlier comments regarding the significant findings and clinical care implications of this study, I must admit that I am a little disappointed that Mann et al used the old medical approach to disability. From a public health perspective, the old medical approach to disability is not the ideal approach to injury research among children with disabilities. In fact, Mann et al did not even mention the “biopsychosocial” approach to disability, which has framed both the new research agenda at the National Institute on Disability and Rehabilitation Research and the 2005 Surgeon General’s Call to Action to Improve the Health and Wellness of Persons with Disabilities (6). The new biopsychosocial approach emphasizes the interaction between impairments, activities, and participation in the physical and social environments. The medical approach, on which most studies have been based in the past (including many of my own publications), emphasizes only the individual’s physical, cognitive, or psychiatric impairments and does not take into account how other aspects of the individual’s life may be affected by his/her impairments. Thus, this biopsychosocial approach to disability may provide a better conceptual framework for future injury research and the development of injury interventions among children with disabilities.

References

1. Mann JR, Zhou L, McKee M, McDermott S. Children with hearing loss and increased risk of injury. Ann Fam Med. 2007;5(6):528-33.

2. Ramirez M, Peek-Asa C, Kraus JF. Disability and risk of school related injuries. Inj Prev. 2004; 10(1): 21-26.

3. Xiang H, Stallones L, Chen G, Hostetler SG, Kelleher K. Nonfatal injuries among U.S. children with disabling conditions. Am J Public Health. 2005;95(11):1970-1975.

4. Sinclair SA, Xiang H. Injuries among U.S. children with different types of disabilities. Am J Public Health (published ahead of print at http://www.ajph.org).

5. Chen G, Sinclair SA, Smith GA, Ranbom L, Xiang H. Incidence and patterns of burn injuries among children with disabilities. J Trauma. 2007;62(3):682-686.

6. U.S. Department of Health and Human Services. The Surgeon General's Call To Action To Improve the Health and Wellness of Persons with Disabilities. Rockville, MD: Office of the Surgeon General, U.S. Department of Health and Human Services; 2005.

Competing interests:   None declared

We would like to thank Dr. Osterling for the insights regarding the relevance of our article for clinical practice. We agree that tailored interventions for preventive education and counseling are an important field for future research. These interventions should address issues related to hearing per se, but also should help provide parents/caregivers with skills for effective hazard communication with children who have hearing loss. Of course, it is vital that any intervention developed be evaluated for effectiveness prior to large scale implementation. Meanwhile, it would seem advisable for clinicians to be aware of the apparent increased risk of injury in these children and to discuss this with parents.

Competing interests:   None declared