Objective. This study examines 1) the way that children with chronic conditions are cared for at home and assisted by technology affects maternal employment and child care; 2) the social and clinical factors associated with the decision of a mother to quit employment to care for a child at home; and 3) the way in which care at home and the decision of a mother to quit a job affects maternal mental health.
Design. The 6-month postdischarge status of 70 mothers of children assisted by technology (study group) was compared with the 6-month postdischarge status of 58 mothers of children (matched for age and gender) hospitalized for acute illnesses (comparison group). Between January and December 1993, we gathered information on sociodemographic status, employment status and changes in employment, severity of the child's condition, child care and nursing services at home, family support, and maternal mental health.
Results. One third of mothers in the study group reported that they quit employment to take care of a child at home with only 37.1% remaining employed outside the home, compared with 69.0% of comparison group mothers. Single caretakers were 15 times more likely to quit employment compared with mothers in two-parent families. Availability of child care had an independent effect on a mother's decision to quit a job, whereas the severity of the child's condition did not. Child care hours were significantly lower in study group families and were provided mostly by relatives compared with day-care facilities and regular babysitters in comparison families. Family support was highest among employed mothers in both the study and the comparison groups and lowest in study group mothers who were neither employed currently nor before the child's illness or who had quit employment to care for the child. Family income was significantly lower in families with a child assisted by technology. Families in the study group had 20-fold higher uncompensated health care costs than did the comparison group. Mothers caring for a child assisted by technology reported less good mental health than did comparison group mothers, and employment seems to mediate this relationship.
Conclusions. Caring for a child assisted by technology seems to create barriers to maternal employment diminishing family resources at a time when financial needs actually may increase. Lack of family support and child care services increase the likelihood that mothers of children assisted by technology will stay out of the labor force. Remaining employed buffers the negative effects of care at home on maternal mental health. Health policies for children with chronic health problems should address issues of financial burdens and the labor force participation of their caretakers. chronic illness, home care, technology assisted, family support, employment, quality of life, child care.
Growing numbers of children who have chronic conditions and who are assisted by technology are cared for at home, most of them by their mothers. This study examines how care at home affects maternal mental health, maternal employment, and child care. We were interested specifically in the social and clinical factors associated with a mother's decision to quit employment to care for a child at home.
Increasing numbers of women with dependent children are employed outside the home. In the United States, labor force participation of mothers with children younger than 6 years of age doubled from 30% to 59% between 1970 and 1990.1 Factors affecting maternal employment include socioeconomic status, education, fertility, household composition, availability of child care and nursing, and other caretaking responsibilities of women for children or the elderly.2 Although all employed parents of young children encounter difficulties with emergencies such as a child's acute illness, the problem exacerbates with a child with a complex medical condition requiring frequent and unpredictable medical care. Although more fathers participate in child care than in previous generations, mothers still carry the major burden of child and child health care for both acute and chronic conditions3 ,4 and may forgo employment opportunities to take care of children.5 ,6Previous studies have found moderately lower rates of employment among mothers of children with chronic conditions,7 ,8 especially among lower-income households.9 ,10 However, most studies took place before the growth of technology, and, therefore focused on chronic conditions in general and at a time when fewer women were employed. One more recent study also included families with children assisted by technology and found lower rates in employment in this group compared with families with children with other chronic conditions.11
Developments in medical care and new technologies for children with complex medical conditions have increased longevity impressively;12 ,13 along with increasingly high costs for hospital care, better ways to finance less costly home care since the mid-1960s14 and increasing evidence of the negative effects of long-term hospitalization on child development have encouraged the move from hospital to home for many of these children. Taking home a child with a complex medical condition, especially when assisted by medical devices, can be a major life event leading to chronic stress in the family.15 ,16 Parents, particularly mothers, must prevent or manage medical crises, administer treatments, restructure family time, and ensure the well-being of all family members. Home health services rarely cover all the child's needs, leaving a major portion of nursing to the child's caregivers.17 ,18 Only a handful of special day care facilities for children assisted by technology are in operation.19 Family support may help to alleviate stress and support family members in coping.20 The ability of the primary caregiver to recruit support and child care resources may be crucial to her participation in the labor force and in the determination of her well-being and quality of life.21 ,22Severity of the child's condition seemed to have an impact on maternal labor force participation in one study of mothers of children with a variety of chronic illnesses.9 No data are available regarding whether this association exists in families with children assisted by technology.
Child health outcomes depend in part on the caregiver's ability to provide adequate health care at home and to ensure psychological well-being and social development. Negative financial experiences, such as lost income when caregivers leave work outside the home to care for an ill child and large out-of-pocket expenses related to the child's health problem, likely affect the way that parents care for their children. Maternal mental health also is associated with health outcomes in children with chronic conditions.23 Employment or the decision to quit employment may affect not only family finances but also maternal mental health24 ,25 and thus the ability of the mother to care for a child assisted by technology at home.
The relationship between care at home of children with chronic conditions assisted by technology and choices in employment opportunities has not been well studied. In this study, we sought to determine the effects of caring for a child assisted by technology on maternal employment and decisions to quit jobs. More specifically, we expected to find less maternal employment in families with a child assisted by technology and that clinical severity of the condition and indicators of family support and child care options would affect maternal employment. We also expected to find that less employment is associated with less good mental health among mothers of children assisted by technology.
We conducted a cross-sectional study of employment of mothers 6 months after the child's initial hospitalization leading to technology assistance. We included a comparison group of families with apparently healthy children matched for age and gender to estimate the effects attributable to the technology needs beyond those shared by comparable young families in the community who also had a child hospitalized. We studied initially whether having a child assisted by technology was associated with decreased maternal employment and then we studied the social and clinical correlates of mothers' employment decisions when they had a child who was assisted by technology.
Children were eligible for the study if they had been hospitalized at the Massachusetts General Hospital, a large tertiary care center in Boston, between January and December 1993 and were discharged from the hospital to their primary caregivers. The study was approved by the Subcommittee on Human Studies, Committee on Research, Massachusetts General Hospital. Children in the study group were identified at the time of hospital discharge using the following criteria: 1) presence of a chronic condition of anticipated duration of ≥6 months and 2) onset of a need for technology assistance as defined by the Office of Technology Assessment13 during this hospitalization. Chronic conditions in the study group included birth defects, prematurity, recurrent apnea, malignancies, cerebral palsy and epilepsy, obstructive intestinal or kidney disease, and severe airway disease.
Children in the comparison group were eligible if they had experienced short hospitalizations for acute conditions within the same sampling period as the children assisted by technology matched for age and gender. They were identified through chart review of hospital discharges for acute diagnoses not indicating chronic illness (transient respiratory problems of the newborn, urinary tract infection without malformations of the renal tract, pyloric stenosis, intussusception, bronchiolitis, croup, cellulitis, lymphadenitits, and appendicitis). Those diagnoses were selected to ensure an age and gender distribution similar to that for the study group. Children in the comparison group were excluded if their charts indicated a chronic illness or technology assistance.
After identification and enrollment of eligible children, we first sought permission from both the primary care physician and the attending physician in the hospital to contact the family. After professional consent, we contacted the family by mail, explaining the study and inviting both parents to participate. All participating parents returned a signed consent form. Families received a written survey 6 months after discharge and were reminded up to three times by telephone if they failed to return the survey within 4 weeks. Both groups received the same survey.
For parental employment, we asked mothers and fathers to indicate whether they were currently employed outside the house. If they were employed, we asked for average number of hours worked per week; if they were not employed, we asked whether they had quit a job to care for a child assisted by technology. We combined the information in three categorical variables: employed, mothers who were employed both before and after the child's hospitalization; homemaker, mothers who were not employed before or after the child's hospitalization; and quit, mothers who had been employed before the child's hospitalization but said that they had quit a job to care for the child. The last category applied only to mothers in the study group. (We acknowledge the fact that mothers in all three groups have responsibilities as homemakers. The variable name homemaker simply implies that these mothers occupied homemaking but not labor force roles before and after the child's hospitalization.)
For additional work-related problems, we asked whether mothers felt that they had to work fewer hours compared with the time before the child's hospitalization; whether they had to take time off when the child was sick; whether they had to take a different job to accommodate child care needs; whether they earned less because they took care of a child with a health problem; and whether they felt unable to change employment for fear of losing health insurance for their child. These questions applied to employed mothers only.
To assess the severity of the child's condition we used the Functional Status II-R developed by Stein and Jessop.26 This measure is comparable with an activity of daily living assessment and probes the child's ability to perform age-appropriate tasks; higher scores indicate higher functioning. The survey is age-specific with three versions: one for children up to 1 year of age; a second for children from 1 to 2 years old; and a third for older children. Scoring for the third version varies according to the child's age. The authors report high internal consistency with α > .80 and good discriminant and construct validity.
We also included several questions related to the child's health care and nursing needs in our surveys. We asked families to indicate which technological device(s) their child used during the previous 6 months, eg, tracheostomy, ventilator, airway suctioning, nebulizer, oxygen, monitor, gastrostomy, enteral feeding tube, ileostomy or colostomy, urostomy, dialysis, clean intermittent catheterization, central venous line, infusion pump, and/or cardiac pacemaker. The number of devices was added to calculate a sum of assistive devices used by each child. To tap the amount of professional services, we asked families whether they received home health nursing or home health aides and the number of nursing or aide hours per week. To evaluate child care services, we asked families to indicate whether they received child care or respite services for the index child from day-care services, regular babysitters, relatives, or other sources and to estimate the hours provided per week by each source. The total of all hours per week was calculated and called child care hours.
We used the Family Resource Index developed by Moos and Moos27 to assess family support. This composite index combines three subscales, cohesion, expressiveness, and conflict, from the Family Environment Scale. Cohesion is conceptualized as the degree of commitment, help, and support among family members; expressiveness is defined as the extent to which family members act openly and express their feelings directly; and conflict is defined as the amount of openly expressed anger, aggression, and conflict among family members. The index has high internal consistency and good construct validity28 and has been used extensively as a summary measure of family support.29 ,30
Sociodemographic information included maternal age and education, child age and gender, family size (total number of adults and children in the household), household characteristics (single parent or not), and ethnicity. Family financial resources included self-reported income in the year before the survey (ie, including the period of the child's hospitalization and the 6 months after) with nine income categories (under $5000; $5000–$9999; $10 000–$19 999; $20 000–$29 999; $30 000–$39 999; $40 000–$59 999; $60 000–$79 999; $80 000–$99 999; $100 000 or more). In addition, we asked families to estimate out-of-pocket expenses related to the child's illness during the previous year in various areas, eg, transportation, medical bills, and extra child care, and added the expenses to obtain total expenses over the year.
Maternal mental health was determined from the SF-36 Health Survey,31 an instrument widely used to assess health-related quality of life. The scale assesses different health concepts including limitations in physical activities, social activities, usual role activities, pain, vitality, and general health perceptions. The instrument has excellent construct and external validity32 with population-based normative values. Factor analysis confirmed the existence of two distinct components: mental and physical health summary scales, accounting for 80% to 85% of the variance in the SF-36 subscales.33 In this report, we present the results from the mental health summary scale.
For bivariate comparison of sociodemographic data and employment status in the two groups, we used frequency tables and χ2tests and for categorical data with cell sizes <10, we used Fisher's exact test. We used t tests to compare means of continuous data with normal distribution and nonparametric tests (Wilcoxon) for continuous data with nonnormal distribution.34 Next, we classified mothers according to their employment status and whether they had a child assisted by technology (study group: employed, homemaker, or quit; comparison group: employed or homemaker). To compare means among the groups, we used ANOVA for continuous data with normal distribution or general linear models approximation for continuous data with nonnormal distribution.
We used logistic regression to determine the effects of severity, sociodemographic factors, nursing and child care, and family support on the mother's choice to quit a job. Independent variables were removed from the model if they did not show an independent effect (P < .1) on the choice and did not change the parameters of the remaining variables by >20% with their removal. To assess the relationship of a continuous, nonnormally distributed outcome (mothers' mental health) with categorical variables (group membership and employment status), we used general linear models approximations that enabled us to test for interaction effects of the independent variables on the dependent variable.35 All statistical procedures were supported by Statistical Analysis Software, Triangle Inc.36
We identified 104 children for the study group and their physicians permitted contact with 98 of the families. Of these, 75 (76.5%) study group parents returned the survey. In the comparison group, we enrolled 131 families and their physicians permitted contact with 125. Of these families, 68 (54.4%) answered the survey. Participants and nonparticipants were similar in child age and gender; nonparticipants in both the study and the comparison groups were more likely to be of African-American or Hispanic ethnicity (P < .001), to live in less affluent communities, and to be insured through Medicaid. Information on employment status was available on 70 mothers and 60 fathers in the study group and 58 mothers and 54 fathers in the comparison group. For participation in the labor force, we report rates of both mothers and fathers. For additional examination of factors associated with employment or changes in employment, we restrict our analysis to mothers.
Sociodemographic characteristics of the study and comparison groups were similar (Table 1) except for maternal employment rates, family income, and out-of-pocket expenses. As expected, children in the study group had less good health status, indicated by lower functional status scores (Table 1). Children in the study group used a wide variety of devices, including tracheostomies, oxygen concentrators, nebulizers, airway suctioning, ventilators, and monitors, feeding tubes, gastrostomies, ileostomies, urostomy, and intermittent catheterization, and central venous lines and infusion pumps. Most study group children used one device, but 31 required more support with up to eight devices at home. There were no significant differences in functional status within the study-group children relevant to maternal employment status.
Employment Situation of Mothers and Fathers and Sociodemographic Factors
Nearly all fathers were employed outside the home (53 [88.3%] in the study group and 50 [92.6%] in the comparison group); one father in the study group had quit a job during the 1rst months after discharge to take care of his child. The employment situations of the mothers were different. During the year after the child's discharge, 26 (37.1%) mothers in the study group were employed, compared with 40 (69.0%) in the comparison group (P < .01). A total of 23 (32.9%) study group mothers stated that they had quit a job to care for their child. Among employed study group mothers, 12 of 26 (46%) reported working fewer hours than they had previously, 6 (23%) had taken a different job to accommodate child care needs, 12 (46%) believed that they earned less because they took care of a child assisted by technology, and 11 (42%) believed that they were unable to change employment because they feared losing health insurance for their child. None of these problems were encountered by families in the comparison group. The need to take time off when the child is sick was mentioned by 24 (92%) of employed mothers in the study group and 14 (35%) in the comparison group (P < .01). Number of work hours per week for employed mothers did not differ significantly (Table 1).
Six months after discharge, family income differed significantly between study and comparison groups, with 35.7% of study group families earning <$30 000 in the previous year, compared with 19.0% in the comparison group (P < .05) (Table 1). Income was associated significantly with employment status (Table 2). In addition to reduced family income, families in the study group reported an average of uncompensated health care costs of $5062 (±8564) for hospital bills, doctor charges, and medications, travel, child care related to doctor visits, diets, special equipment, home modifications, and other expenses related to the child's illness. Expenses for the comparison group families were $265 (±473).
Factors Affecting Maternal Employment
For both groups, maternal education was associated significantly with employment status: 38 (57.6%) of all mothers employed outside the home had some college education, whereas only 9 (23.1%) of 39 homemakers had some college education. Study group mothers who had quit a job to take care of a child were similar to the employed mothers; 52.2% had a college education (Table 2). Mothers who had quit a job had larger families than employed mothers (P < .05). Minority status was significantly associated with the likelihood to be a homemaker but did not differ between the study and comparison groups. A total of 12 (17.1%) study group mothers were single parents, compared with 7 (12.1%) comparison group mothers. Although Table 2shows no significant differences across all five subgroups, when employed mothers were compared with homemakers, single parents and minorities were less likely to be employed outside the home (P < .05).
The logistic regression results indicate that comparison group mothers were greater than three times as likely to work outside the home than were mothers with children assisted by technology (Table 3). Although other sociodemographic variables (maternal education, minority status, single parent, and child age) were associated independently with maternal employment, adding them to the regression did not change the group effect. Mothers with a college education were 2.5 times more likely to be employed; mothers who were not minority were 3.5 times more likely to be employed. The association of single-parent status and employment lost significance when maternal education and ethnicity were taken into account in the same regression (Table 3).
Nursing Care, Child Care, and Family Support
Approximately half of all families in both groups (55.7% in study group and 58.7% in comparison group) received help for child care from day care services, regular babysitters, or relatives.
Families with children assisted by technology relied more heavily on relatives and fewer on day care facilities. Although no differences existed in the age distribution of children, only 1 child assisted by technology but 12 comparison-group children were in day care. Seven (10.0%) study group children had regular babysitters compared with 42 comparison group children (71.2%); however, 32 study-group children were cared for by relatives (46.4%) compared with 18 comparison-group children (31.6%) (Table 4). Parents were least likely to have child care when children used more than one assistive device. In addition to child care hours, 25 study-group children (35.7%) received professional support by home nurses or health aides.
Mothers employed outside the home received significantly more child care hours (sum of hours per week provided by day care, babysitters, and relatives) than mothers at home: employed study group mothers received 12.5 hours; study group homemakers received 6.4 hours; and those mothers who quit a job received 4.6 hours. Comparison group mothers employed outside the home received 18.2 hours compared with the 4.2 hours that homemakers received (Table 5). The number of nursing hours per week did not differ among mothers who were employed, not employed, and who had quit a job. Adding the number of child care hours to those from professional services, families in the study group received a total of 18.0 hours per week with 14.2 hours in the comparison group (P < .01).
Mothers in families with children assisted by technology experienced significantly less family support (on the Family Resources Index) than mothers with apparently healthy children (140.2 ± 27.9 vs 954.4 ± 19.0; P < .01); however, employment status was associated independently with family support across groups with homemakers generally experiencing lower family support than mothers employed outside the home (P < .05) (Table 5).
Factors Affecting the Choice of the Mother to Quit Employment
We compared the 26 mothers in the study group who remained employed with the 23 who had quit a job by 6 months after the child's discharge to assess factors associated with the choice of the mother to quit a job to care for a child assisted by technology. Bivariate analyses indicated significant associations of single parent status, less available child care hours, and less family support with decisions to quit employment. In logistic regression, both single parent status and child care hours remained significant. Single mothers were almost 15 times more likely to quit a job compared with mothers living in two-parent families. Use of child care also contributed significantly to the decision of a mother to quit employment with the odds to quit a job increasing by 10% for every child care hour less (Table 6). Family support lost significance in multivariate analysis with single parent status in the same regression (Wald χ2 1.31;P = .25) and was removed from the model. Severity of the child's condition, home health service use, and other sociodemographic factors did not affect the mother's choice to quit a job.
Maternal Mental Health and Employment
Mothers with a child assisted by technology reported a mean mental health score of 41.6, compared with a mean of 50.3 in comparison group mothers (higher values indicate better mental health) (Table 1). Maternal mental health also was associated with employment status (Table 2). Among study group mothers, those employed outside the home reported the highest scores for mental health, compared with mothers who had quit a job and mothers not employed currently or before the child's hospitalization. Among comparison group mothers, we found no significant differences in mental health relative to employment status. General linear model analysis with maternal mental health scores as the dependent variable and group membership, employment status, and the interaction of group and employment as independent variables showed a significant interaction effect (P < .01). Employment in the study group but not the comparison group seems to act as a protective factor with relatively high levels of psychological well-being in employed study group mothers compared with study group mothers who were not employed. The primary effect of caring for a child assisted by technology on maternal mental health remained significant (P < .001), but primary effect of employment (P < .1) did not.
In the current study, about one third of mothers of a child assisted by technology reported quitting a job to care for the child at home. This decision affected half of all mothers in the study group who were employed outside the home before the child's illness and/or hospitalization. If they had not quit jobs, the study and comparison groups would have had similar rates of employment among mothers. Those who remained employed reported significant work-related problems including having to work fewer hours, to take time off, to change jobs to accommodate care at home, or to remain at jobs because they feared the loss of health insurance. Having a child assisted by technology particularly affected maternal employment among single parents. Single mothers with a child assisted by technology had much higher rates of quitting a job than did mothers in two-parent families. Although single mothers in general encounter hardships that two-parent families may not necessarily face,37 the need for ongoing supervision and care for a child with a chronic condition and the lack of substitute care make the situation even more difficult.38
Breslau and others found an employment rate of 43% compared with 48% in control families with apparently healthy children in the 1970s (when maternal employment generally was less common),39 with a stronger effect of child disability on maternal labor force participation in low-income and black families than among high-income and white families.9 Contrary to the results of our study, increasing severity contributed to the likelihood of no employment or fewer work hours. In a more recent British study, 42% of mothers taking care of a young adult with severe handicapping condition participated in the labor force (with less than one third working full-time), compared with 75% of mothers in the comparison group.7 Of the mothers not employed, 82% attributed their nonemployed status to limitations arising from the disabilities of their child. There was no relationship between the degree of disability and employment.
Quitting a job likely had a major financial impact. We have no data on family income before the child's hospitalization; however, the comparison group provides some pertinent information. Insofar as the group had similar sociodemographic status, based on the parents' age, education, ethnicity, and family composition, it seems likely that study group families initially had incomes comparable with comparison group families and suffered substantial losses in income over the year. In a study of 120 children hospitalized for injury in Massachusetts, 16.7% of parents said they had to quit a job to take the child home, and almost all described serious financial problems for families with more severe injury associated with highest risk for long-term sequelae.40 Families with children with chronic conditions report substantial out-of-pocket expenses in this and other studies.41 In addition to extra expenses not covered by health insurance, lost income may contribute to stress for families caring for an ill child.
Families with and without children assisted by technology had similar rates of child care hours; parents of children assisted by technology had more difficulty recruiting regular substitute care, especially day care. Not all day care centers accept children with disabilities.42 ,43 A recent study of New Haven area child care found that most (65%) centers would accept chronically ill and handicapped children, yet the low enrollment in these centers suggests that other barriers exist.44 Special centers for children assisted by technology exist in only very few places across the country.22 In this study, relatives had a higher share in supporting families with children assisted by technology, although for some families skilled home nursing services provided their only source of respite. Skilled nursing did not seem like surplus hours for parents of children assisted by technology but like a substitute for child care hours that most other parents enjoy but these families are unable to recruit.
The family support findings demonstrated a complex pattern. Although family support was lower in the study group, mothers not employed account for all the difference. Homemakers and mothers who had quit a job reported very low levels of family support. With our cross-sectional data, we do not know whether low levels of family support contributed to or followed a mother's decisions to quit a job. The finding that mothers who may most need family support receive the least is a concern. Our measure of family support, the Family Resource Index, combines dimensions of family cohesion, expressiveness, and conflict. Caretaking could affect all three dimensions: family members may vary in their commitment to care at home or resent the time demands on the mother; family members may have less ability to express their feelings freely or openly express anger, aggression, and conflict.
The mental health of the mother was less good in those mothers who were caring for a child assisted by technology compared with comparison group mothers. The multivariate analyses suggest that employment had a protective effect on mental health in mothers caring for children assisted by technology. Previous studies have indicated that employment among mothers of children with chronic conditions was associated with less depression, controlling for socioeconomic status and child condition.45 Home and work environments may exert complex effects on maternal mental health.25 For example, mothers with rewarding jobs seem to be protected from the negative mental health effects of troubled relationships at home.46 Under stressful circumstances at home, the work environment and a break from domestic routine responsibilities may help mothers to cope with caring for a child with a serious chronic health condition.
Our study suggests that having a child assisted by technology may force many mothers to quit employment, diminishing family resources at a time when financial needs may actually increase. These findings support the need for strengthening several family resources, including social and other supportive services, and providing financial assistance to replace lost income. The Supplemental Security Income program for children and adolescents provides such financial aid to low-income households with children with severe mental, developmental, or physical abilities. The use of health care social workers experienced in coordination of care as well as counseling may help families making decisions concerning care at home and out-of-home work schedules and may aid family adaptation.47 Employment outside the home not only may be important economically, but may have implications for the mental health of mothers.
The data also indicate the importance of improving nursing and child care services for this population, recognizing the complex nature of their health conditions. Lack of respite care, day care, or after-school programs for children assisted by technology seemed to be a major problem for the families in this study. Community-based programs should aim at the inclusion of children with chronic conditions and those assisted by technology. These services may particularly benefit mothers who remain employed. In addition, certain labor policies, such as flexible time arrangements or additional sick leave to take care of a sick family member, may also help to alleviate stress in families with children assisted by technology.48 ,49
This study was supported by grants from Deutscher Akademischer Austauschdienst and Sonderprogramm Epidemiologie, Bonn, Germany, and the Deborah Munroe Noonan Memorial Fund, Boston, Massachussetts.
We thank Elizabeth Anderson from the Federation for Children with Special Needs, Boston, MA, for her invaluable support of the project.
- Received September 15, 1998.
- Accepted December 17, 1998.
Reprint requests to (U.T.) Klinik für Pädiatrie, Medizinischen Universität zu Lübeck, Kahlhorststr 31–35, 23538 Lübeck, Germany. E-mail:
- ↵US Bureau of the Census. Statistical Abstracts of the United States. Washington, DC: US Government Printing Office; 1992
- American Academy of Pediatrics
- Carpenter ES
- ↵Baldwin S. The Cost of Caring: Families with Disabled Children. London, UK: Routledge and Kegan; 1985
- Raddish M,
- Bloom SR,
- Gortmaker SL,
- Perrin JM
- ↵US Congress. Office of Technology Assessment. Technology-Dependent Children: Hospital vs. Home-care. Washington, DC: Government Printing Office; 1987. Technical Memorandum OTA-TM-H-38
- ↵Patterson JM, Leonhard BJ, Titus JC. Home care for medically fragile children: impact on family health and well-being. J Dev Behav Pediatr. 1992;248–255
- ↵Fleming J, Challela M, Eland J, et al. Impact on the family of children who are technology dependent and cared for in the home. Pediatr Nurs. 1994;379–388
- ↵Aday LA, Wegener DH, Andersen RM, Aitken MJ. Home care for ventilator assisted children. Health Aff. 1989; Summer:137–147
- ↵Perrin JM, Shayne MW, Bloom SR. Home and Community Care for Chronically Ill Children. New York, NY: Oxford University Press; 1993
- ↵Thompson RJ, Gustafson KE. Adaptation to Chronic Childhood Illness. Washington, DC: American Psychological Association; 1996
- ↵Stein REK, Jessop DJ. Functional Status II (R). A measure of child health status. Med Care. 1990;1041–1055
- ↵Moos R, Moos B. Family Environment Scale Manual. Palo Alto, CA: Consulting Psychologists Press; 1994
- ↵Holahan CJ, Moos RH. The quality of social support. Measures of family and work relationships. Br J Clin Psychol. 1983;157–162
- ↵Billings A, Moos R. Social support and functioning among community and clinical groups: a panel model. J Behav Med. 1982;295–311
- ↵Frey KS, Greenberg MT, Fewell RR. Stress and coping among parents of handicapped children: a multidimensional approach. Am J Ment Retard. 1989;240–249
- ↵Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health Survey Manual and Interpretation Guide. Boston, MA: The Health Institute, New England Medical Center; 1993
- ↵Ware JE, Kosinski M, Keller SD. SF-36 Physical and Mental Health Summary Scales: A User's Manual. Boston, MA: The Health Institute, New England Medical Center; 1994
- ↵Pagano M, Gauvreau K. Principles of Biostatistics. Belmont, CA: Duxbury Press; 1992
- ↵Kleinbaum DG, Lawrence LK, Muller KE. Applied Regression Analysis and Other Multivariate Methods. 2nd ed. Belmont, CA: Duxbury Press; 1987
- ↵SAS/STAT Software: Release 6.10. Carey, NC: SAS Institute Inc; 1994
- Osberg JS,
- Kahn P,
- Rowe K,
- Brooke MM
- Heyman SJ,
- Earle A,
- Egleston B
- Copyright © 1999 American Academy of Pediatrics