PEDIATRICS Vol. 119 No. 2 February 2007, pp. e391-e398 (doi:10.1542/peds.2006-1459)
ARTICLE |
Guardianship Planning Among HIV-Infected Parents in the United States: Results From a Nationally Representative Sample
a Department of Pediatrics, Mattel Children’s Hospital, David Geffen School of Medicine
b Department of Health Services, School of Public Health, University of California, Los Angeles, California
c RAND, Santa Monica, California
d Department of Psychology, Virginia Commonwealth University, Richmond, Virginia
e Los Angeles County Department of Public Health, Los Angeles, California
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
OBJECTIVE. The purpose of this work was to determine the rates and predictors of guardianship planning and preferred guardians among HIV-infected parents.
PARTICIPANTS AND METHODS. Data were analyzed from interviews with 222 unmarried parents (who had 391 children) from a nationally representative sample of HIV-infected adults receiving health care. Outcome measures included parental report on the level of guardianship planning and on who their preferred guardian for each child was. Level of guardianship planning was categorized as follows: (1) parent had not identified a guardian; (2) parent had identified a guardian, but the guardian had not agreed; (3) identified guardian had agreed; and (4) legal documentation of guardianship plan was complete. We conducted bivariate and ordered logistic regression analyses on the level of guardianship planning and multinomial logistic regression on identification of preferred guardians.
RESULTS. Twelve percent of unmarried HIV-infected parents had not identified a guardian; 6% had identified a guardian but gone no further; 53% said the identified guardian had agreed; and 28% had prepared legal documentation. The preferred guardians included other biological parents (17%), spouse/partners who were not biological parents (2%), grandparents (36%), other relatives (34%), friends (7%), unrelated adoptions (1%), and others (3%). Parents with the lowest CD4 counts and parents living without other adults were more likely to have completed the guardianship planning process. Nonrelatives were most often preferred by mothers and parents with higher CD4 counts; grandparents were most often preferred by younger parents and parents who prefer speaking Spanish.
CONCLUSIONS. Pediatricians and others who take care of children with HIV-infected parents may be able to provide counseling and referrals for guardianship planning.
Key Words: HIV • children and adolescents
Abbreviations: HCSUS—HIV Cost and Services Utilization Study • MSA—metropolitan statistical area • HAART—highly active antiretroviral therapy
HIV-infected parents express concerns about planning for the future care of their children in the event of their own death.1–3 When parents arrange for their child's guardianship in advance of their own death, children often cope better.4,5 Specifically, guardianship planning can decrease the likelihood of children being shuffled from one home to another, spending prolonged periods in foster care, or being separated from siblings.4 Children who lose a parent to HIV have higher rates of depression, sexual risk behaviors,6 and other emotional and behavioral problems.7 The few empirical studies on the adjustment of children orphaned by AIDS highlight the importance of providing children with a supportive family environment and limited living transitions after a parent's death.8,9 Guardianship planning can temper the potential effects of parental death on the children. Physicians and others treating children of HIV-infected parents may be able to refer parents to resources that facilitate the guardianship planning process.
Planning for the future care of children can be a difficult process for an HIV-infected parent. HIV-related stigma, cultural beliefs, depression, fear of mortality, and distrust and misunderstanding of the legal system can create challenges that make it hard for HIV-infected parents to make legal arrangements for their children's care.1,2,10–13 In addition, some parents may have trouble identifying a suitable guardian11 or may assume that a family member will become their child's guardian in the absence of a legal document. A study of HIV-infected parents in New York, NY, found that 54% changed their guardianship plans over 5 years, and at death, only 57% of parents had a legal guardianship plan in the form of a will, standby guardianship arrangement, or both.14
We have only a rudimentary understanding of whether parents make formal guardianship plans and why many do not. Previous studies have involved focus groups or interviews with convenience samples of mostly HIV-infected mothers receiving care at an HIV clinic in a single large city.9,14–17 These studies found that mothers (compared with fathers), parents with younger children, and parents diagnosed with HIV longer ago were more likely to have legal documentation of their guardianship plan. Parental age, race or ethnicity, and physical health status were not related to guardianship planning. Although these studies have provided useful information, nationally representative data on this topic have not been available. Therefore, we used data from a nationally representative sample of HIV-infected adults receiving health care to examine how far in the guardianship planning process parents had progressed and whom they had identified as the preferred guardian. Our study is also the first to include a multilevel outcome of guardianship planning.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study Design
Respondents participated in the HIV Cost and Services Utilization Study (HCSUS), which selected a national probability sample of people
18 years old with known HIV infection who made 1 visit to a nonmilitary, nonprison medical provider other than an emergency department in the contiguous United States during the first 2 months of 1996. This article draws on 2 waves of data collected from January 1996 to April 1997 (baseline) and from December 1996 to June 1997 (follow-up). The main outcome variables in this article come from the follow-up survey. Details of the HCSUS design and methods appear elsewhere.18–20
For this article, our analysis sample includes all of the unmarried parents who participated in both survey waves and who had custody of 1 of their children at follow-up. Married parents were not included, because the surviving spouse would often retain legal custody of the children, and a formal guardianship document would usually be unnecessary. We did not have information on whether the spouse was the biological or adoptive parent of the child. Our analysis consisted of 222 unmarried parents with 391 children (Fig 1).
|
Trained interviewers used computer-assisted personal interviewing for in-person interviews in English or Spanish lasting
90 minutes.21 The RAND Institutional Review Board and local boards approved the study.
Measures
Predictor Variables
Variables collected at baseline included the following: parent's and child's birth date (used to calculate age at time of interview), gender, educational attainment, household income, HIV risk/exposure group, year of HIV diagnosis, geographic region (ie, Midwest, Northeast, South, or West), and metropolitan statistical area (MSA) (rural to 1.5 million, 1.5–2.5 million, 2.5–4.5 million, or >4.5 million). Parents reported their race/ethnicity (black, Latino/Hispanic, white/other). White/other included Asian/Pacific Islander, Indian/Alaskan Native, and other, which were too few to analyze separately.
At baseline and follow-up, respondents reported on use of highly active antiretroviral therapy (HAART) during the 6 months before the interview. A 3-category variable was constructed to indicate whether respondents had initiated HAART at baseline, at follow-up, or never.
At follow-up, respondents reported their lowest CD4 count. If parents did not know the exact CD4 count, they were asked if their lowest count was 500/mL, 200 to 499/mL, 50 to 199/mL, or <50/mL. High levels of agreement have been found between self-report and medical chart CD4 counts in a hospitalized population.22 Respondents also reported their number of overnight hospital stays during the previous 6 months. The living situation for unmarried parents at follow-up was classified as living with or not living with other adults. Latino/Hispanic respondents were asked 5 questions about their English/Spanish use; other respondents were assumed to prefer English.
Respondents were asked at both survey waves about the HIV status of each of their children (
5 children). Children who were not reported to be HIV positive were categorized as HIV negative/unknown status.
Outcome Variables (Follow-up Survey)
Parents were asked to report on these items separately for 5 children (aged 0–17 years) who were in the parent’s custody.
Level of Guardianship Planning
This 4-level ordinal variable described the highest level of planning and was derived from 4 items: (1) "guardian not identified" indicated parents who responded "no" to, "if anything were to happen to you, do you know who you would like to have custody of this child?"; (2) "identified a guardian" indicated parents who responded "yes" to that question but not to additional questions; (3) "guardian agreed" indicated parents who responded "yes" to, "has this person agreed to be this child's legal guardian?" but not to the final question; and (4) "legal document prepared" indicated parents who responded "yes" to, "do you have a will or legal document that says this person will be this child's legal guardian if you were to die or become incapacitated?"
Preferred Guardian
Respondents who had identified a guardian were asked, "what is this person's relationship to this child?" Answer options were as follows: other biological parent, spouse/partner who is not the biological parent, grandparent, other relative, friend, unrelated adoptive or foster parent, or another person. For a multivariate multinomial logistic regression predicting selection of guardians other than biological parents or spouse/partners, the variable was collapsed into 3 categories: grandparents, other relatives, and friends/other nonrelatives.
Data Analysis
For the ordinal level of guardianship planning outcome, we conducted bivariate and multivariate ordered logistic regression analyses using the child as the unit of analysis. For the multivariate ordered logistic regression model, we included predictors for which bivariate analyses on the overall sample had 2-sided P values of <.20, a standard screening threshold.23 This threshold is less stringent than the P <.05 used to evaluate statistical significance so as to avoid inappropriately removing predictors that have stronger multivariate than bivariate effects. Parent age, educational attainment, household income, HIV risk/exposure group, hospital stays, year of HIV diagnosis, MSA, and geographic region exceeded the screening threshold and were not included in the multivariate analysis. All of the remaining variables were parameterized categorically in the final multivariate model, as shown in Table 1.
|
A second ordered logistic regression was conducted on the 95 parents (with 259 children) who reported on >1 child. This ordered logistic regression predicted the level of guardianship planning from child age after controlling for family level fixed effects. For all of the multivariate ordered logistic regressions, the assumption of proportionate odds that underlies ordered logistic regression was assessed using the Brandt test.24 We report adjusted odds ratios resulting from ordered logistic regression.
For children (n = 275) whose parent identified a preferred guardian other than a biological parent or spouse/partner, we used a multinomial logistic regression to investigate whether the choice of preferred guardian varied by parent characteristics. We conducted the analysis on this subset because the preference for a biological parent or spouse/partner would generally be dominated by whether one existed.
The statistical power of the main ordinal logistic regression, given its distribution, is roughly equivalent to that of an ordinary least squares (OLS) regression with sample size 158 but without the assumption of linear scaling among guardianship categories.25 To avoid overfitting, we limited ourselves to 14 predictor degrees of freedom, within the guidelines of 1 per 10 observations in ordinary least squares regression.26 Considering all of the design effects, we had 80% power with 2-sided tests to detect a "moderate" effect size.27 It should be noted that small associations between candidate predictors and the outcomes may have gone undetected. Power would have been sufficient to detect only very large interactions among predictors, so interactions were not investigated.
We report weighted percentages for categorical outcomes, both overall and by covariates. We also report whether each category of a given variable differed significantly from the corresponding omitted category with a Wald t test. To limit multiple testing, we performed the Wald t test only with categorical variables for which there was a statistically significant overall effect of the variable as a whole in the multivariate ordinal logistic regression (eg, whether the ordinal outcome varied overall by race/ethnicity, as determined by joint postestimation within the ordinal logistic regression).
The analytic weights take into account differential selection probabilities: nonresponse, multiplicity, and attrition.28 The weights also adjust for a computer error that caused 5% of eligible parents to not be administered the parenting questions. All of the analyses use modifications of HCSUS weights that incorporate the number of children within families and account for this and other aspects of the complex sample design, including the clustering of children within families, using Stata survey commands.29
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Characteristics
Table 1 shows weighted characteristics of unmarried parents and their children.
Level of Guardianship Planning
Twelve percent of parents had not identified a guardian for their children, 6% had only identified a guardian, 53% said the guardian had agreed (but without documentation), and 28% had prepared legal documentation (Table 2).
|
Sixteen percent of parents with >1 child had levels of guardianship planning that differed among their children (data not in Table 1). An ordered logistic regression predicting the level of guardianship planning for children in families with multiple children found no significant effect of child age after controlling for family level fixed effects (P = .510); in other words, we found no tendency for guardianship plans to be more fully developed for older or younger children within a family.
Predictors for Level of Guardianship Planning
In bivariate analysis (Table 2), lower CD4 count is associated with more complete guardianship planning (P = .006). For example, 57% of parents with these CD4 counts had prepared legal documentation, as opposed to 20% to 30% for other CD4 categories. The guardianship planning process is also more complete for children with parents who lived without other adults in the household compared with parents who lived with other adults (P = .025). Parents who prefer speaking English were further along in the planning process than those who prefer speaking Spanish (P = .047). The guardianship planning process is more complete for HIV-positive children compared with HIV-negative/unknown-status children (P = .024).
In multivariate analysis, the findings for CD4 counts and living with other adults persisted, but Spanish language preference and child HIV status were no longer significant (Table 2). In particular, for any given planning threshold, living with no other adults was associated with almost twice the odds of having achieved that planning threshold, and having a CD4 count <50/mL was associated with 3 to 5 times the odds of having achieved that threshold when compared with any other CD4 count. For a sensitivity analysis, we conducted an ordered logistic regression excluding children whose parent selected the other biological parent as the preferred guardian; there were no substantial changes.
Because the Brant test showed a violation of the assumption of proportional odds, we used unweighted generalized ordered logistic regression as a sensitivity test.24,30 The conclusions reached by that approach were quite similar to those reported here, with the exception that the generalized ordered logistic regression suggested that the effects of living without other adults were strongest at the earlier stages of planning and that the effects of CD4 count varied slightly across the levels of planning.
Preferred Guardian
Among parents who identified a guardian, 17% listed the other biological parent as the preferred guardian, followed by a spouse/partner who was not the biological parent (2%), grandparent (36%), another relative (34%), a friend (7%), an unrelated adoptive or foster parent (1%), and other (3%; Table 3).
|
For parents selecting a guardian other than the biological parent or spouse/partner who was not the biological parent, choice of guardian varied significantly by parent characteristics (P < .001 overall for the multivariate multinomial regression). Fathers were less likely than mothers to designate nonrelatives (P < .001), and parents with higher CD4 counts were more likely than parents with lower CD4 counts (P = .008) to designate nonrelatives. Older parents were more likely than younger parents to choose other relatives (P < .001) or nonrelatives (P = .003) compared with grandparents. Parents who prefer speaking English versus Spanish were more likely to choose other relatives (P = .009) or nonrelatives (P = .001) compared with grandparents.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among unmarried parents in a nationally representative sample of HIV-infected adults receiving health care, only 28% of their children had a legally documented guardianship plan, leaving
75% at risk for an unstable transition after their parent's death. Guardianship planning is believed to help children cope with the loss of a parent,4,5 perhaps by reducing the likelihood that a child is shuffled from one home to another, spends prolonged periods in foster care, or is separated from siblings.4 Adolescent children of HIV-infected parents may also experience increased depressive symptoms and sexual risk behaviors soon after their parent's death.6 By assuring that a formal guardianship plan is in place, parents can prevent their children from experiencing added hardships at an already difficult time. Many HIV-infected parents who have identified a guardian for their children are not formalizing their plans. Legal options for guardianship planning generally include a will or standby guardianship.31 A standby guardianship enables parents to petition the court to appoint a standby guardian who can become a temporary guardian if the parent is incapacitated or a permanent guardian if the parent dies.32
More than half of the parents in our study had identified a guardian who agreed to care for the children but had not completed legal documentation. An additional 6% of the parents identified a guardian but had not discussed their wishes with the individual. These parents almost always preferred grandparents or other relatives as guardians (when not selecting the other biological parent); they may have assumed that relatives or family friends would be willing to take the child and that no one would challenge such a plan. Even when a relative is the obvious person to take a child, the lack of legal documentation can slow the process and leave children subject to an uncertain future. For example, without a formal guardianship plan, caregivers may not have the authority to enroll children in school or provide consent for medical procedures.33 Physicians and other clinicians may be able to ask HIV-infected parents if they have a guardianship plan in place and either provide advice on the subject or, if the clinician is not adequately informed, refer them to knowledgeable clinical staff or appropriate social service organizations for assistance.
A parent's choice of guardian can also have important implications for the guardianship planning process. In our study, fathers and parents with higher CD4 counts were more likely to designate nonrelatives as the preferred guardian. In these instances, a formal guardianship plan would be necessary to guarantee that the children were placed with the intended guardian, especially if a relative might contest the plan. In addition, younger parents and parents who prefer speaking English were more likely to choose the children's grandparents as the guardian. In these families, the parent should not assume that the grandparent would automatically receive guardianship of the children on their death, but this desire should be recorded in a formal guardianship plan. Similarly, a parent's preference for a spouse/partner who is not a biological parent (and who has not adopted the child) to serve as guardian should be formally documented, because biological relatives may otherwise contest this arrangement.
Parents' ability or perceived need to select a guardian for their children and document their choice may depend on factors pertaining to family characteristics and/or social support. In our study, 57% of unmarried parents lived with another adult, and these parents were less likely to have completed a formal guardianship plan compared with parents who lived alone. This finding supports the hypothesis that parents who are surrounded by family or friends may not feel the need to formalize their plans. In black communities, children are often cared for during a time of family crisis by relatives and close family friends.34,35 In Latino communities, extended support networks of family and friends frequently assist with child rearing.36 In these and other communities, informal support systems may provide a parent with a sense of comfort, knowing there will be someone to care for their children. Unfortunately, the presence of such a support system may lead parents to not make formal guardianship plans. Clinicians and other professionals may want to consider family factors, such as household composition and child-rearing approach, when discussing guardianship planning with HIV-infected parents.
Formal guardianship arrangements may not be made for other reasons. Parents may have difficulty identifying suitable guardians to care for their children when they do not have family or friends available to assume this role.11 As mentioned above, parents may also fail to complete guardianship plans because of HIV-related stigma, depression, and distrust of the legal system.1,2,10 Clinicians and social service organizations may need to be sensitive to such issues.
The influence of an HIV-infected parent's illness severity on their propensity to complete a formal guardianship plan has not been thoroughly explored in past studies. One study conducted in New York City found no association between an increasing score on a physical health symptom scale and guardianship planning.17 In our study, parents with the lowest CD4 counts were more likely than parents with a CD4 count in one of the higher categories to have completed a guardianship plan. As parents' health declines, they may feel it is more important to formalize a plan for their children's future. Physicians and others treating HIV-infected parents may want to advise parents or refer them to appropriate support services before they experience worsening health status.
Many children in the United States have lost 1 or both parents to HIV. One study estimated that >97000 children had lost their mother to HIV/AIDS as of 1998.37 The introduction of HAART in the mid-1990s slowed the progression of HIV disease.38,39 At the time of our study, HAART had recently become available, and fewer HIV-infected adults were taking HAART compared with today. HAART may have been new enough that it was not influencing people's planning behaviors, but now that HAART is more common, we would expect that parents would be less likely to plan because of longer life expectancies. However, averages may be deceptive. Some parents will become incapacitated or die sooner than expected, so guardianship planning remains important.
Although we had access to a rich set of predictors, there is the possibility, as with all regression analyses of observational data, that unmeasured confounding variables may have influenced the results. Here such potential confounders might have included the health and HIV status of potential guardians. Previous studies have found a significant relationship between child age and the presence of a formal guardianship plan.14,16,17 Child age was not significant in our study, perhaps because our representative sample reflects broader national patterns rather than local variations.
Our study may be relevant to parents with other serious illnesses. Little has been written about guardianship planning for parents with other potentially fatal illnesses. A study of oncology patients who were single parents found that for 40% of them, the children ultimately went to people whom the deceased parents would have opposed.40 If a parent dies without a formal plan in place, courts may place children with family members or friends who were already living with and taking care of the children, but this cannot be assured if formal documentation is not present, especially if the other biological parent wishes to regain custody of the children.41 Advice books for parents with a serious illness provide guidance to parents with HIV/AIDS, cancer, and other illnesses,42,43 but few studies have covered the topic. Parents are faced with the often-challenging need to discuss their guardianship plans with their children and the preferred guardians. Physicians may want to address such issues not only with HIV-infected parents but with any parents with a serious illness, if not all parents.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by National Institute of Child Health and Human Development grant R01 HD40103 and Centers for Disease Control and Prevention grant U48/DP000056. The original data collection was supported in part by Agency for Health Care Policy and Research grant U01HS08578.
We are indebted to Jacinta Elijah, BA, Theresa Nguyen, BS, and Jennifer Patch, BA, for research assistance. We also thank the HCSUS Consortium for making the study possible and the study participants for sharing their time and stories.
|
FOOTNOTES |
|---|
Accepted Aug 14, 2006.
Address correspondence to Burton O. Cowgill, MPH, Department of Pediatrics, School of Medicine, University of California, 1072 Gayley Ave, Los Angeles, CA 90024. E-mail: bcowgill{at}ucla.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
This work does not necessarily represent the opinions of the funding organizations or of the institutions with which the authors are affiliated.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Tompkins TL, Henker B, Whalen CK, Axelrod J, Comer LK. Motherhood in the context of HIV infection: reading between the numbers. Cultur Divers Ethnic Minor Psychol. 1999;5 :197 –208[CrossRef]
- Bauman LJ, Draimin BH, Levine C, Hudis J. Who will care for me? Planning the future care and custody of children orphaned by HIV/AIDS. In: Pequegnat W, Szapocznik J, eds. Working With Families in the Era of HIV/AIDS Thousand Oaks, CA: Sage Publications, Inc; 2000:155 –189
- Draimin BH, Gamble I, Shire A, Hudis J. Improving permanency planning in families with HIV disease. Child Welfare. 1998;77 :180 –194[Web of Science][Medline]
- Gardner W, Preator K. Children of seropositive mothers in the U.S. AIDS epidemic. J Soc Issues. 1996;52 :177 –195[Web of Science]
- Dane BO, Levine C, eds. AIDS and the New Orphans: Coping With Death Westport, CT: Auburn House/Greenwood Publishing Group; 1994
- Rotheram-Borus MJ, Weiss R, Alber S, Lester P. Adolescent adjustment before and after HIV-related parental death. J Consult Clin Psychol. 2005;73 :221 –228[CrossRef][Web of Science][Medline]
- Pelton J, Forehand R. Orphans of the AIDS epidemic: an examination of clinical level problems of children. J Am Acad Child Adolesc Psychiatry. 2005;44 :585 –591[CrossRef][Web of Science][Medline]
- Rotheram-Borus MJ, Leonard NR, Lightfoot M, Franzke LH, Tottenham N, Lee S-J. Picking up the pieces: caregivers of adolescents bereaved by parental AIDS. Clin Child Psychol Psychiatry. 2002;7 :115 –124[CrossRef]
- Forehand R, Pelton J, Chance M, et al. Orphans of the AIDS epidemic in the United States: transition-related characteristics and psychosocial adjustment at 6 months after mother's death. AIDS Care. 1999;11 :715 –722[CrossRef][Web of Science][Medline]
- Armistead L, Forehand R. For whom the bell tolls: parenting decisions and challenges faced by mothers who are HIV seropositive. Clin Psychol Sci Prac. 1995;2 :239 –250[Web of Science]
- Mason S. Custody planning with HIV-affected families: considerations for child welfare workers. Child Welfare. 1998;77 :161 –177[Web of Science][Medline]
- Faithful J HIV-positive and AIDS-infected women: challenges to mothering. Am J Orthopsychiatry. 1997;67 :144 –151[Web of Science][Medline]
- Ingram D, Hutchinson SA. Defensive mothering in HIV-positive mothers.
Qual Health Res. 1999;9
:243
–258
[Abstract/Free Full Text] - Rotheram-Borus MJ, Lester P, Wang PW, Shen Q. Custody plans among parents living with human immunodeficiency virus infection.
Arch Pediatr Adolesc Med. 2004;158
:327
–332
[Abstract/Free Full Text] - Simoni JM, Davis ML, Drossman JA, Weinberg BA. Mothers with HIV/AIDS and their children: disclosure and guardianship issues. Women Health. 2000;31 :39 –54[CrossRef][Web of Science][Medline]
- Rotheram-Borus MJ, Draimin BH, Reid HM, Murphy DA. The impact of illness disclosure and custody plans on adolescents whose parents live with AIDS. AIDS. 1997;11 :1159 –1164[CrossRef][Web of Science][Medline]
- Lightfoot M, Rotheram-Borus MJ. Predictors of child custody plans for children whose parents are living with AIDS in New York City. Soc Work. 2004;49 :461 –468[Web of Science][Medline]
- Shapiro MF, Berk ML, Berry SH, et al. National probability samples in studies of low-prevalence diseases. Part I: perspectives and lessons from the HIV cost and services utilization study. Health Serv Res. 1999;34 :951 –968[Web of Science][Medline]
- Frankel MR, Shapiro MF, Duan N, et al. National probability samples in studies of low-prevalence diseases. Part II: designing and implementing the HIV cost and services utilization study sample. Health Serv Res. 1999;34 :969 –992[Web of Science][Medline]
- Schuster MA, Kanouse DE, Morton SC, et al. HIV-infected parents and their children in the United States.
Am J Public Health. 2000;90
:1074
–1081
[Abstract/Free Full Text] - Berry SH, Brown JA, Athey LA, et al. HCSUS Baseline Patient Questionnaire Documentation Santa Monica, CA: RAND; 1998
- Cunningham WE, Rana HM, Shapiro MF, Hays RD. Reliability and validity of self-report CD4 counts-in persons hospitalized with HIV disease. J Clin Epidemiol. 1997;50 :829 –835[CrossRef][Web of Science][Medline]
- Hosmer DW, Lemeshow S. Applied Logistic Regression New York, NY: Wiley; 1989
- Brandt R. Assessing Proportionality in the proportional odds model for ordinal logistic regression. Biometrics. 1990;46 :1171 –1178[CrossRef][Web of Science][Medline]
- Taylor A, West S, Aiken L. Loss of power in logistic, ordinal logistic, and probit regression when an outcome variable is coarsely categorized. Educ Psychol Meas. 2006;66
:228
–239
[Abstract/Free Full Text] - Harrell F, Lee K, Mark D. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15 :361 –387[CrossRef][Web of Science][Medline]
- Cohen J. Statistical Power Analysis for the Behavioral Sciences Hillsdale, NJ: Erlbaum; 1988
- Duan N, McCaffrey DF, Frankel MR, et al. HCSUS Baseline Methods Technical Report: Weighting, Imputation, and Variance Estimation. Santa Monica, CA: RAND; 1999. MR-1060-AHCPR
- Stata Statistical Software [computer software]. Release 7.0. College Station, TX: Stata Corporation; 2001
- Williams R. Generalized ordered logit/partial proportional odds models for ordinal dependent variables. Stata J. 2006;6 :58 –82
- Hudis F, Brown J. Custody and permanency planning. In: Joslin D, ed. Invisible Caregivers: Older Adults Raising Children in the Wake of HIV/AIDS. New York, NY: Columbia University Press; 2002
- Herb A. Standby Guardianship: a viable legal option for the future care of children. J Am Med Womens Assoc. 1995;50 :95 –98[Medline]
- Coon L LJ, Manes C, Palmer C. Guide to Future Care and Custody Planning for Children: With Recommendations for State Legislation Berkeley, CA: National Abandoned Infants Assistance Resource Center, University of California; 2005
- Brown S CD, Wheeler R. African American extended families and kinship care: how relevant is the foster care model to kinship care. Child Youth Serv Rev. 2002;24 :53 –77[CrossRef][Web of Science]
- Chatters LTR, Jayakody R. Fictive kinship relations in black extended families. J Comp Fam Stud. 1994;25 :297 –313[Web of Science]
- Vidal C. Godparenting among Hispanic Americans. Child Welfare. 1988;67 :453 –459[Web of Science]
- Lee LM, Fleming PL. Estimated number of children left motherless by AIDS in the United States, 1978–1998. J Acquir Immune Defic Syndr. 2003;34 :231 –236[CrossRef][Web of Science][Medline]
- Gange SJ, Barron Y, Greenblatt RM, et al. Effectiveness of highly active antiretroviral therapy among HIV-1 infected women.
J Epidemiol Community Health. 2002;56
:153
–159
[Abstract/Free Full Text] - Louie JK HL, Osmond DH, Katz MH, Schwarcz. Trends in causes of death among persons with acquired immunodeficiency syndrome in the era of highly active antiretroviral therapy, San Francisco, 1994–1998. J Infect Dis. 2002;186 :1023 –1027[CrossRef][Web of Science][Medline]
- Willis L, Peck M, Sells S, Rodabaugh KJ. Custody planning. A retrospective review of oncology patients who were single parents. J Pain Symptom Manage. 2001;21 :380 –384[CrossRef][Web of Science][Medline]
- Cooper E. HIV-infected parents and the law: issues of custody, visitation, and guardianship. In: Hunter N, Rubenstein W, eds. AIDS Agenda: Emerging Issues in Civil Rights. New York, NY: New Press; 1992
- McCue K. How to Help Children Through a Parent's Serious Illness New York, NY: St Martin's Griffin; 1996
- Goldman L. Breaking the Silence: A Guide to Help Children With Complicated Grief: Suicide, Homicide, AIDS, Violence, and Abuse New York, NY: Brunner-Routledge; 2001
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
