Advertisement

A Prospective Study of Sudden Unexpected Infant Death after Reported Maltreatment

Published:October 17, 2013DOI:https://doi.org/10.1016/j.jpeds.2013.08.073

      Objective

      To examine whether infants reported for maltreatment face a heightened risk of sudden infant death syndrome (SIDS) and other leading causes of sudden unexpected infant death (SUID).

      Study design

      Linked birth and infant death records for all children born in California between 1999 and 2006 were matched to administrative child protection data. Infants were prospectively followed from birth through death or 1 year of age. A report of maltreatment was modeled as a time-varying covariate; risk factors at birth were included as baseline covariates. Multivariable competing risk survival models were used to estimate the adjusted relative hazard of postneonatal SIDS and other SUID.

      Results

      A previous maltreatment report emerged as a significant predictor of SIDS and other SUID. After adjusting for baseline risk factors, the rate of SIDS was more than 3 times as great among infants reported for possible maltreatment (hazard ratio: 3.22; 95% CI: 2.66, 3.89).

      Conclusion

      Infants reported to child protective services have a heightened risk of SIDS and other SUID. Targeted services and improved communication between child protective services and the pediatric health care community may enhance infant well-being and reduce risk of death.
      ASSB (Accidental suffocation and strangulation in bed), CPS (Child protective services), HR (Hazard ratio), SIDS (Sudden infant death syndrome), SUID (Sudden unexpected infant death)
      Each year in the US, approximately 4500 children die during the first 12 months of life with no immediately identifiable cause or explanation, occurrences broadly defined as sudden unexpected infant death (SUID).
      • Kinney H.C.
      • Thach B.T.
      The sudden infant death syndrome.

      Sudden unexpected infant death and sudden infant death syndrome. Centers for Disease Control and Prevention Web site. Available at: http://www.cdc.gov/sids. Updated September 4, 2012. Accessed December 31, 2012.

      • Shapiro-Mendoza C.K.
      • Camperlengo L.T.
      • Kim S.Y.
      • Covington T.
      The Sudden Unexpected Infant Death Case Registry: a method to improve surveillance.
      More than one-half of SUIDs are ultimately classified as caused by sudden infant death syndrome (SIDS),
      • Shapiro-Mendoza C.K.
      • Kimball M.
      • Tomashek K.M.
      • Anderson R.N.
      • Blanding S.
      US infant mortality trends attributable to accidental suffocation and strangulation in bed from 1984 through 2004: are rates increasing?.
      the designation for infant deaths that remain unexplained following a thorough autopsy, death scene investigation, and review of the child's medical history.
      • Willinger M.
      • James L.S.
      • Catz C.
      Defining the sudden infant death syndrome (SIDS): deliberations of an expert panel convened by the National Institute of Child Health and Human Development.
      Another 14% of these initially unexplained deaths are coded as accidental suffocation and strangulation in bed (ASSB), and approximately 30% are certified as deaths of unspecified cause.
      • Kim S.Y.
      • Shapiro-Mendoza C.K.
      • Chu S.Y.
      • Camperlengo L.T.
      • Anderson R.N.
      Differentiating cause-of-death terminology for deaths coded as sudden infant death syndrome, accidental suffocation, and unknown cause: an investigation using US death certificates, 2003-2004.
      The current distribution of SUID classifications reflects a diagnostic shift largely attributed to increasing medical examiner and coroner adherence to criteria for certifying a death as SIDS.
      • Willinger M.
      • James L.S.
      • Catz C.
      Defining the sudden infant death syndrome (SIDS): deliberations of an expert panel convened by the National Institute of Child Health and Human Development.
      Deaths that would have once been coded as SIDS may now be certified as ASSB or unspecified in cause.
      • Shapiro-Mendoza C.K.
      • Kimball M.
      • Tomashek K.M.
      • Anderson R.N.
      • Blanding S.
      US infant mortality trends attributable to accidental suffocation and strangulation in bed from 1984 through 2004: are rates increasing?.
      • Kim S.Y.
      • Shapiro-Mendoza C.K.
      • Chu S.Y.
      • Camperlengo L.T.
      • Anderson R.N.
      Differentiating cause-of-death terminology for deaths coded as sudden infant death syndrome, accidental suffocation, and unknown cause: an investigation using US death certificates, 2003-2004.
      • Shapiro-Mendoza C.K.
      • Tomashek K.M.
      • Anderson R.N.
      • Wingo J.
      Recent national trends in sudden, unexpected infant deaths: more evidence supporting a change in classification or reporting.
      Of concern is that regardless of how deaths are coded, postneonatal mortality rates have remained static since 2001.
      • Task Force on Sudden Infant Death Syndrome
      The changing concept of sudden infant death syndrome: diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk.
      • Task Force on Sudden Infant Death Syndrome
      SIDS and other sleep-related infant deaths: expansion of recommendations for a safe infant sleeping environment.
      The American Academy of Pediatrics expanded its policy on safe-sleeping practices in 2005 with specific recommendations for sleep surfaces, bedding, separate but proximate sleeping environments, and use of pacifiers.
      • Task Force on Sudden Infant Death Syndrome
      The changing concept of sudden infant death syndrome: diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk.
      Yet in 2009, the US mortality rate for SIDS was 53.9 per 100 000 live births—nearly unchanged from prior years and consistent with a similarly flat rate of postneonatal deaths.

      Kochanek KD, Xu J, Murphy SL, Miniño AM, Hsiang-Ching K. Deaths: final data for 2009. Available at: http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_03.pdf. Published December 29, 2011. Accessed December 31, 2012.

      The absence of continued declines in the rate of SIDS led to a further expansion of sleep-related recommendations in 2011, the impact of which is yet unknown.
      • Task Force on Sudden Infant Death Syndrome
      SIDS and other sleep-related infant deaths: expansion of recommendations for a safe infant sleeping environment.
      A child's previous involvement with child protective services (CPS) has been documented as a risk factor for preventable death.
      • Putnam-Hornstein E.
      Report of maltreatment as a risk factor for injury death: a prospective birth cohort study.
      • Jonson-Reid M.
      • Chance T.
      • Drake B.
      Risk of death among children reported for nonfatal maltreatment.
      A similar relationship between CPS contact and SIDS, however, has not emerged. An early study found that the prevalence of CPS investigations was not higher among 157 cases of SIDS compared with a similar group of living infants.
      • O'Halloran R.L.
      • Ferratta F.
      • Harris M.
      • Ilbeigi P.
      • Rom C.D.
      Child abuse reports in families with sudden infant death syndrome.
      A study by Krous et al
      • Krous H.F.
      • Haas E.A.
      • Manning J.M.
      • Deeds A.
      • Silva P.D.
      • Chadwick A.E.
      • et al.
      Child protective services referrals in cases of sudden infant death: a 10-year, population-based analysis in San Diego County, California.
      examined 384 cases of SIDS over a 10-year period and found that an earlier substantiated report of maltreatment was not associated with an increased likelihood of SIDS. Krous et al
      • Krous H.F.
      • Haas E.A.
      • Manning J.M.
      • Deeds A.
      • Silva P.D.
      • Chadwick A.E.
      • et al.
      Child protective services referrals in cases of sudden infant death: a 10-year, population-based analysis in San Diego County, California.
      noted the limitations of a retrospective study and suggested future research that would allow for comparisons to living infants.
      This study builds upon prior research by using linked birth, CPS, and death records to prospectively model the relative likelihood of a SUID following a report of maltreatment. We conceptualized a report to CPS, regardless of substantiation, as a signal of a high-risk infant likely to experience poorer outcomes and greater adversities. We hypothesized that after adjusting for other risk factors, infants reported for possible maltreatment would face a heightened risk of SIDS and other forms of SUID.

      Methods

      To create the analysis dataset, linked birth and infant death records were matched to CPS data. Confidential birth and infant death records were obtained from the California Department of Public Health. Files captured all children born in California between 1999 and 2006 and included death record information for infants who died before 1 year of age. Infant CPS records were available from the California Department of Social Services through a longstanding interagency agreement with the University of California, Berkeley. This project received approval from both state and university committees for the protection of human subjects.
      To link vital records to CPS data, a probabilistic matching methodology was employed in which a record pair was deemed a match/nonmatch based on a formal statistical model.
      • National Program of Cancer Registries
      Link Plus User's Guide, Version 2.0.
      Match-status cut-points were established following an extensive examination of record pairs, with a clerical review of uncertain matches falling between established lower- and upper-bound match thresholds.
      • Herzog T.N.
      • Scheuren F.J.
      • Winkler W.E.
      Data Quality and Record Linkage Techniques.
      Across birth cohorts, 90% of CPS records were matched to a birth record. Unmatched records reflected infants born in other states and subsequently reported for maltreatment in California, as well as records in which missing data elements prevented a successful match. Further details regarding the linkages underlying this dataset have been previously published.
      • Putnam-Hornstein E.
      Report of maltreatment as a risk factor for injury death: a prospective birth cohort study.
      Our final analytic dataset consisted of the full population of children born in California between 1999 and 2006 (N = 4 317 736), with corresponding CPS and death records through each child's first birthday. From this population of live births, we excluded 15 402 children who did not survive the neonatal period (ie, died before 28 days of life). We also excluded 599 children with dates of birth/death that were either missing or did not logically align.

      Variables

       Dependent Variable

      Postneonatal deaths were classified using the International Classification of Diseases, 10th Revision.
      • World Health Organization
      International statistical classification of diseases and related health problems: Tenth revision.
      SUID were defined as deaths coded as SIDS (code R95), unspecified cause (code R99), or ASSB (code W75).
      • Shapiro-Mendoza C.K.
      • Tomashek K.M.
      • Anderson R.N.
      • Wingo J.
      Recent national trends in sudden, unexpected infant deaths: more evidence supporting a change in classification or reporting.
      For each model, all other causes of postneonatal deaths were coded as competing, censored events.

       Independent Variable

      We examined an earlier, nonfatal report of maltreatment as an independent predictor of SUID. We included all reports, including those screened out over the phone prior to an investigation, as well as those that were investigated and determined to be unfounded. This decision is consistent with a growing body of research suggesting that a report to CPS is a meaningful marker of risk, regardless of whether or not the report is substantiated.
      • Hussey J.M.
      • Marshall J.M.
      • English D.J.
      • Knight E.D.
      • Lau A.S.
      • Dubowitz H.
      • et al.
      Defining maltreatment according to substantiation: distinction without a difference?.
      • Thompson R.
      • Wiley T.R.
      Predictors of re-referral to child protective services: a longitudinal follow-up of an urban cohort maltreated as infants.
      A nonfatal report was entered into our models as a time-varying covariate based on the report date. As such, infants remained in the general population of infants born and at risk of death until the date they were first reported, at which point they were prospectively followed as an infant with an earlier report to CPS.

       Other Covariates

      Risk factors for SIDS/SUID have been identified in prior research. Higher rates of SIDS have been observed among infants who are male, African American, or Native American.
      • Malloy M.H.
      • MacDorman M.
      Changes in the classification of sudden unexpected infant deaths: United States, 1992-2001.
      • Overpeck M.D.
      • Brenner R.A.
      • Cosgrove C.
      • Trumble A.C.
      • Kochanek K.
      • MacDorman M.
      National underascertainment of sudden unexpected infant deaths associated with deaths of unknown cause.
      • Haglund B.
      • Cnattingius S.
      Cigarette smoking as a risk factor for sudden infant death syndrome: a population-based study.
      • Trachtenberg F.L.
      • Haas E.A.
      • Kinney H.C.
      • Stanley C.
      • Krous H.F.
      Risk factor changes for sudden infant death syndrome after initiation of Back-to-Sleep campaign.
      Developmental risk factors include premature birth, low birth weight, and 2-4 months of age.
      • Kinney H.C.
      • Thach B.T.
      The sudden infant death syndrome.
      • Overpeck M.D.
      • Brenner R.A.
      • Cosgrove C.
      • Trumble A.C.
      • Kochanek K.
      • MacDorman M.
      National underascertainment of sudden unexpected infant deaths associated with deaths of unknown cause.
      • Trachtenberg F.L.
      • Haas E.A.
      • Kinney H.C.
      • Stanley C.
      • Krous H.F.
      Risk factor changes for sudden infant death syndrome after initiation of Back-to-Sleep campaign.
      • Ostfeld B.M.
      • Esposito L.
      • Perl H.
      • Hegyi T.
      Concurrent risks in sudden infant death syndrome.
      Family/maternal risk factors for SIDS include young maternal age, short intergestational interval, single motherhood, late or absent prenatal care, and low socioeconomic status.
      • Kinney H.C.
      • Thach B.T.
      The sudden infant death syndrome.
      • Overpeck M.D.
      • Brenner R.A.
      • Cosgrove C.
      • Trumble A.C.
      • Kochanek K.
      • MacDorman M.
      National underascertainment of sudden unexpected infant deaths associated with deaths of unknown cause.
      • Haglund B.
      • Cnattingius S.
      Cigarette smoking as a risk factor for sudden infant death syndrome: a population-based study.
      Additional risks for SIDS include an unsafe sleep environment,
      • Schnitzer P.G.
      • Covington T.M.
      • Dykstra H.K.
      Sudden unexpected infant deaths: sleep environment and circumstances.
      as well as gestational cigarette and/or alcohol exposure.
      • Haglund B.
      • Cnattingius S.
      Cigarette smoking as a risk factor for sudden infant death syndrome: a population-based study.
      • Wisborg K.
      • Kesmodel U.
      • Henriksen T.B.
      • Olsen S.F.
      • Secher N.J.
      A prospective study of smoking during pregnancy and SIDS.
      • Milerad J.
      • Vege A.
      • Opdal S.H.
      • Rognum T.O.
      Objective measurements of nicotine exposure in victims of sudden infant death syndrome and in other unexpected child deaths.
      In the present study, adjustments were made for 7 covariates available in the birth records: (1) child's sex (male vs female); (2) maternal race/ethnicity (Black, Latino, Asian/Pacific Islander, or Native American vs White); (3) birth weight (<2500 g vs ≥2500 g); (4) prenatal care (first trimester care vs late or absent care); (5) birth payment method (public insurance vs private insurance); (6) maternal age (≤19 years vs >19 years); and (7) paternity (missing vs established). In addition, our modeling approach captured exposure time from birth to death, and therefore incorporated infant age.

       Statistical Analyses

      All infants were prospectively followed from birth through death or 1 year of age. The distribution of sociodemographic, health, and developmental characteristics were assessed using χ2 tests. Comparisons were made between (1) SUID (deaths coded as R95, R99, or W75) and infants surviving to 1 year of age; and (2) SUID classified as SIDS (R95) vs those that were unspecified (R99) or coded as ASSB (W75). Multivariable competing risk survival models were used to estimate the hazard of a sudden unexpected death following an infant's report to CPS as a possible victim of maltreatment.
      • Cleves M.
      • Gutierrez R.G.
      • Gould W.
      • Marchenko Y.V.
      An Introduction to Survival Analysis Using Stata.
      • Fine J.P.
      • Gray R.J.
      A proportional hazards model for the subdistribution of a competing risk.
      This modeling technique accounted for the fact that each infant was at risk of not only SIDS, ASSB, and death coded as unspecified, but also other postneonatal causes of death. Observations were censored upon death or 1 year of age. Model estimates are reported as hazard ratios (HR) and 95% CI. Statistical analyses were conducted using StataSE (Stata Corp, College Station, Texas).
      • StataCorp
      Stata Statistical Software: Release 12.1.

      Results

      Table I compares the sociodemographic and health characteristics of infants surviving to 1 year of age (n = 4 299 642) to the population of postneonatal SUID (n = 2093). Among infants in the SUID group, 16.6% had been reported to CPS, compared with only 4.9% of infants surviving to 1 year of age (P < .001). Consistent with well-established risk factors for SUID, there were statistically significant differences in the distribution of all covariates examined (P < .001). White, Black, and Native American infants were over-represented among cases of SUID, whereas Hispanic/Latino infants and Asian/Pacific Islander children were under-represented. The proportion of infants with low birth weight was more than 2.5 times greater among infants experiencing SUID than those surviving to 1 year of age, and the proportion of infants born to mothers whose prenatal care started after the first trimester or not at all was more than twice as great. Infants whose birth expenses were covered by the state's public health insurance system were over-represented among cases of SUID.
      Table IDemographics of live births and postneonatal SUID
      Living infants
      Infants surviving until 1 y of age.
      SUID
      Postneonatal SUID defined as a death occurring between 28 and 365 d of age and coded as R95, R99, or W75.
      χ2 TestSIDS (R95)Unspecified (R99)ASSB (W75)χ2 Test
      (n = 4 299 642)(n = 2093)(n = 1263)(n = 703)(n = 127)
      %n%P valuen%n%n%P value
      Previous CPS report
       No report95.1174583.4<.001107184.856880.810683.5.074
       Reported4.934816.619215.213519.22116.5
      Sex
       Female48.984040.1<.00149038.830142.84938.6.205
       Male51.1125359.977361.240257.27861.4
      Race/ethnicity
       White31.177237.1<.00149839.721831.25644.1<.001
       Black6.037618.121617.213819.82217.3
       Latino50.773435.340732.428741.14031.5
       Asian/Pacific Islander11.71617.71128.9446.353.9
       Native American0.5381.8231.8111.643.2
      Birth weight
       <2500 g6.236217.4<.00121517.112818.31915.1.614
       ≥2500 g93.8171982.6104282.957081.710784.9
      Start of prenatal care
       First trimester85.9144070.8<.00187871.747568.98771.9.419
       Late or no care14.159429.234628.321431.13428.1
      Maternal age
       ≤19 y9.842920.6<.00125320.115121.62519.8.706
       >19 y90.2165479.4100679.954778.410180.2
      Insurance coverage
       Public43.5124560.5<.00170456.646167.08064.0<.001
       Private56.581339.554143.522733.04536.0
      Paternity
       Established90.7167780.1<.001103381.854777.89776.4.058
       Missing9.341619.923018.215622.23023.6
      Infants surviving until 1 y of age.
      Postneonatal SUID defined as a death occurring between 28 and 365 d of age and coded as R95, R99, or W75.
      In contrast to the stark risk differences between surviving infants and those experiencing SUID, few differences were observed when those in the SUID group were stratified based on coded cause of death (SIDS, unspecified, or ASSB). There were no statistically significant differences in the prevalence of infants previously reported to CPS across these 3 causes of SUID, nor were there distinctions by child's sex, birth weight, prenatal care, maternal age, or paternity. Only 2 differences across coded causes of death emerged. Proportionately larger shares of SUID occurring among White infants were classified as SIDS or ASSB rather than unspecified, whereas the percentage of Black and Hispanic/Latino infants whose death was classified as unspecified was greater than observed for other causes (P < .001). The cause of SUID among privately insured infants was more likely to be coded as SIDS and less likely to be classified as unspecified (P < .001).

       Infant Age

      Among the approximately 209 000 infants reported as potential victims of maltreatment in the birth cohorts examined, 35.8% (n = 74 966) were reported during the first month of life (Figure 1). The distribution of CPS reports received after the first month was largely static. Among infants experiencing postneonatal SUID (Figure 2), the age distribution was heavily right-skewed for all 3 causes of death, although ASSB fatalities occurring between months 3 and 8 were more uniformly allocated than SIDS and unspecified deaths, which fell steeply.
      Figure thumbnail gr1
      Figure 1Age distribution of first reported maltreatment among infants born in California between 1999 and 2006 and reported to CPS before 1 year of age.
      Figure thumbnail gr2
      Figure 2Age distribution of postneonatal SUID among infants born in California between 1999 and 2006 (28-365 days of age).

       Multivariable Models

      Table II presents results from 4 multivariable competing risk models. These models compared the relative hazard of SUID (Model 1), SIDS (Model 2), unspecified cause (Model 3), or ASSB (Model 4) vs survival to 1 year of age. Each model examined an infant's earlier report to CPS as a prospective risk factor for death, after adjusting for established sociodemographic and health factors associated with both a SUID and a maltreatment report. Maltreatment was entered into the model as a time-varying covariate; all other variables were coded as stable markers of risk as captured at birth.
      Table IIMultivariable competing risk survival models for the leading causes of postneonatal SUID (adjusted HR and 95% CI)
      Observations censored upon death or the child's first birthday, a CPS report was entered into each model as a time-varying covariate. All models adjusted for CPS report, infant sex, infant birth weight, prenatal care, maternal age, maternal race/ethnicity, insurance coverage, and paternity establishment.
      Model 1: SUID (R95, R99, W75)Model 2: SIDS (R95)Model 3: Unspecified (R99)Model 4: ASSB (W75)
      vs survival to 1 y of agevs survival to 1 y of agevs survival to 1 y of agevs survival to 1 y of age
      Adj. HR(95% CI)Adj. HR(95% CI)Adj. HR(95% CI)Adj. HR(95% CI)
      Previous CPS report
       No reportRef.-Ref.-Ref.-Ref.-
       Reported3.49(3.02, 4.02)3.22(2.66, 3.89)4.21(3.32, 5.33)2.30(1.21, 4.36)
      Sex
       FemaleRef.-Ref.-Ref.-Ref.-
       Male1.45(1.32, 1.58)1.52(1.35, 1.71)1.30(1.11, 1.51)1.60(1.11, 2.32)
      Race/ethnicity
       WhiteRef.-Ref.-Ref.-Ref.-
       Black1.55(1.35, 1.77)1.47(1.24, 1.75)1.85(1.46, 2.33)0.97(0.54, 1.74)
       Latino0.43(0.38, 0.48)0.38(0.32, 0.44)0.57(0.47, 0.70)0.27(0.17, 0.45)
       Asian/Pacific Islander0.60(0.50, 0.71)0.63(0.51, 0.78)0.62(0.44, 0.85)0.26(0.10, 0.64)
       Native American1.83(1.29, 2.59)1.86(1.20, 2.89)1.55(0.79, 3.05)2.83(0.98, 8.16)
      Birth Weight
       <2500 g2.46(2.18, 2.78)2.45(2.10, 2.86)2.54(2.07, 3.11)1.94(1.11, 3.39)
       ≥2500 gRef.-Ref.-Ref.-Ref.-
      Start of prenatal care
       First trimesterRef.-Ref.-Ref.-Ref.-
       Late or no care1.75(1.57, 1.95)1.77(1.54, 2.03)1.74(1.46, 2.08)1.61(1.06, 2.46)
      Maternal Age
       ≤19 y1.88(1.68, 2.11)1.94(1.67, 2.25)1.81(1.49, 2.19)1.82(1.14, 2.90)
       >19 yRef.-Ref.-Ref.-Ref.-
      Insurance coverage
       Public1.88(1.68, 2.10)1.68(1.46, 1.94)2.22(1.82, 2.71)2.43(1.51, 3.91)
       PrivateRef.-Ref.-Ref.-Ref.-
      Paternity
       EstablishedRef.-Ref.-Ref.-Ref.-
       Missing1.04(0.91, 1.19)0.99(0.83, 1.17)1.09(0.88, 1.36)1.34(0.80, 2.23)
      Observations censored upon death or the child's first birthday, a CPS report was entered into each model as a time-varying covariate. All models adjusted for CPS report, infant sex, infant birth weight, prenatal care, maternal age, maternal race/ethnicity, insurance coverage, and paternity establishment.
      Among infants who had been reported to CPS as possible victims of abuse or neglect, the rate of SUID was more than 3 times as great (HR: 3.49; 95% CI: 3.02, 4.02). The strong association between a previous report and SUID remained when deaths were stratified by classification as SIDS (HR: 3.22; 95% CI: 2.66, 3.89), unspecified cause (HR: 4.21; 95% CI: 3.32, 5.33), or ASSB (HR: 2.30; 95% CI: 1.21, 4.36). Other covariates—including male sex, race/ethnicity, low birth weight, late prenatal care, public health insurance, and a teenage mother—also emerged as statistically significant predictors of SUID, with only modest differences across death classifications. A notable exception was paternity establishment, which was not significant in any model.

      Discussion

      Although prior studies have explored CPS involvement as a possible antecedent of SIDS,
      • O'Halloran R.L.
      • Ferratta F.
      • Harris M.
      • Ilbeigi P.
      • Rom C.D.
      Child abuse reports in families with sudden infant death syndrome.
      • Krous H.F.
      • Haas E.A.
      • Manning J.M.
      • Deeds A.
      • Silva P.D.
      • Chadwick A.E.
      • et al.
      Child protective services referrals in cases of sudden infant death: a 10-year, population-based analysis in San Diego County, California.
      in this study an earlier report of abuse or neglect was identified as a marker for SIDS and other forms of SUID. We found that after adjusting for health and sociodemographic risk factors—including maternal age, infant birth weight, and prenatal care—infants previously reported as alleged victims of abuse or neglect had a rate of SUID that was more than 3 times higher than infants not reported.
      The emergence of a significantly heightened rate of SIDS and SUID among infants previously reported to CPS lends itself to several interpretations. First, it may be that despite our attempt to adjust for factors that have a demonstrated association with SIDS, infants reported to CPS have unique risks that account for the relationship observed. For example, data suggest that a majority of children reported to CPS come from families in which substance abuse is present,
      • Young N.K.
      • Boles S.M.
      • Otero C.
      Parental substance use disorders and child maltreatment: overlap, gaps, and opportunities.
      and prior research suggests a potential association between prenatal alcohol/drug exposure and SIDS
      • Phillips D.P.
      • Brewer K.M.
      • Wadensweiler P.
      Alcohol as a risk factor for sudden infant death syndrome (SIDS).
      • Kandall S.R.
      • Gaines J.
      Maternal substance use and subsequent sudden infant death syndrome (SIDS) in offspring.
      • Durand D.J.
      • Espinoza A.M.
      • Nickerson B.G.
      Association between prenatal cocaine exposure and sudden infant death syndrome.
      • Iyasu S.
      • Randall L.L.
      • Welty T.K.
      • Hsia J.
      • Kinney H.C.
      • Mandell F.
      • et al.
      Risk factors for sudden infant death syndrome among Northern Plains Indians.
      ; yet, we were unable to adjust for maternal substance use as it is not collected in birth record data.
      A second (and related) interpretation is that the population of infants reported to CPS is composed of a very high-risk subset of infants born into families in which there remains a partial or lagged penetration of public health campaigns. Families of infants reported to CPS may have a concentration of both chronic and acute stressors that reduce adherence to safe-sleeping guidelines. As such, systematic differences between reported and nonreported infants in terms of safe-sleeping habits and other developmentally appropriate protections may explain the heightened rates of SUID. These possible differences in unmeasured environmental factors, coupled with potential differences in unobserved family risk factors related to substance use, may leave infants reported to CPS with a heightened vulnerability to SIDS/SUID.
      A third interpretation is that a previous report for maltreatment is a risk factor because we cannot unequivocally differentiate SIDS from infant deaths caused by soft suffocation, whether accidental or inflicted. It is presently impossible to determine what percentage of SUID classified as SIDS, unspecified cause, or ASSB are related to maltreatment. Although empirical reviews suggest that unascertained infanticide occurs in less than 5% of SIDS cases,
      • Kinney H.C.
      • Thach B.T.
      The sudden infant death syndrome.
      • Hymel K.P.
      National Association of Medical Examiners. Distinguishing sudden infant death syndrome from child abuse fatalities.
      this estimate is based on limited data.
      The emergence of statistically significant differences in SIDS/SUID by sex, maternal age, race/ethnicity, birth weight, and other sociodemographic and health factors aligns with a large body of prior epidemiologic evidence. There were, however, 2 findings that diverged from the literature. First, there was no relationship between SUID and paternity establishment, a surrogate marker for a single mother. Although this is inconsistent with other examinations of single parenthood and SIDS,
      • Blair P.S.
      • Sidebotham P.
      • Berry P.J.
      • Evans M.
      • Fleming P.J.
      Major epidemiological changes in sudden infant death syndrome: a 20-year population-based study in the UK.
      • Leach C.E.
      • Blair P.S.
      • Fleming P.J.
      • Smith I.J.
      • Platt M.W.
      • Berry P.J.
      • et al.
      Epidemiology of SIDS and explained sudden infant deaths.
      these earlier studies defined single status based on whether or not there was a current partner. The present study relied on the establishment of paternity at birth and did not determine whether there was a second caregiver in the months that followed.
      Second, the bivariate distribution of race/ethnicity and health insurance varied across the 3 forms of SUID. Among deaths coded as SIDS and ASSB, a larger proportion of White infants were observed, whereas Black and Latino infants were over-represented among deaths classified as unspecified. Meanwhile, infants covered by public health insurance—a proxy for low socioeconomic status—were over-represented among SUID coded as unspecified and under-represented among cases of SIDS. These variations were notable given the absence of bivariate differences for other risk factors. Although speculative, it may be that these disparities reflect socioeconomic variations in local resources available to investigate infant deaths, a propensity among coroners to assign a cause of death to White and privately insured infant fatalities, or both.
      The interpretation of our data and strength of our conclusions are limited by several factors. First, available birth record data from California did not provide information on several important factors known to increase the risk of SIDS, including fetal and postnatal alcohol and cigarette smoke exposure.
      • Trachtenberg F.L.
      • Haas E.A.
      • Kinney H.C.
      • Stanley C.
      • Krous H.F.
      Risk factor changes for sudden infant death syndrome after initiation of Back-to-Sleep campaign.
      • Wisborg K.
      • Kesmodel U.
      • Henriksen T.B.
      • Olsen S.F.
      • Secher N.J.
      A prospective study of smoking during pregnancy and SIDS.
      • Blair P.S.
      • Sidebotham P.
      • Berry P.J.
      • Evans M.
      • Fleming P.J.
      Major epidemiological changes in sudden infant death syndrome: a 20-year population-based study in the UK.
      Relatedly, although a notable strength of this analysis is its population-based and prospective (rather than retrospective) construction, this design prevented us from incorporating information on known environmental risk factors, such as bed sharing and sleep position, as this information is not available for surviving infants.
      • Schnitzer P.G.
      • Covington T.M.
      • Dykstra H.K.
      Sudden unexpected infant deaths: sleep environment and circumstances.
      Second, current imprecision regarding the measurement of asphyxial risk at the death scene and limited diagnostic methods at autopsy do not allow precise differentiation of SIDS from unspecified causes of death. As a result, these diagnostic labels may not have been consistently applied by death certifiers across various jurisdictions or during the study interval. Finally, the use of California data restricts the generalizability of our findings. Maltreatment reporting rates vary widely across states reflecting different definitions of abuse and neglect and varying community thresholds for reporting an infant to CPS.

      US Department of Health and Human Services. Child maltreatment: 2011. Administration for Children and Families, Administration on Children, Youth, and Families, Children's Bureau Web site. Available at: http://www.acf.hhs.gov/sites/default/files/cb/cm11.pdf. Published December 12, 2012. Accessed December 31, 2012.

      Although our findings were robust to an alternative specification of prior CPS involvement defined as substantiation (findings not reported), we do not know if a similar relationship would emerge in other states.
      In conclusion, data from the past decade suggest there may be a limit to the influence and reach of SIDS awareness and prevention campaigns. If this is the case, efforts to further reduce postneonatal infant deaths will require more intensive interventions tailored and targeted to populations with a greater concentration of risk factors. Although these data fall short of explaining the observed relationship, a previous report to CPS emerged as a prospective risk factor for SIDS and other SUIDs. Although only a small share of infants reported to CPS died from SIDS or SUID, reported infants had a significantly greater rate of death than did other sociodemographically similar children. Likewise, even though only 1 in 6 SUID victims had been reported earlier to CPS, this rate of prior reports was more than 3 times the population prevalence observed among surviving infants. Findings from this study not only underscore the heightened vulnerability of infants reported for maltreatment, but point to opportunities for CPS workers to reinforce safe-sleeping messages, and for the pediatric health care community to engage and monitor families of high-risk infants reported for maltreatment. Further studies of the relationship between SIDS/SUID and a family's earlier involvement with CPS, especially with respect to services offered, are necessary.
      The authors wish to thank Barbara Needell (California Child Welfare Indicators Project at University of California, Berkeley) for her helpful editorial feedback and assistance in data acquisition. The authors also acknowledge colleagues at the California Child Welfare Indicators Project and the California Department of Social Services who contributed to the management and preparation of data underlying this analysis.

      References

        • Kinney H.C.
        • Thach B.T.
        The sudden infant death syndrome.
        N Engl J Med. 2009; 361: 795-805
      1. Sudden unexpected infant death and sudden infant death syndrome. Centers for Disease Control and Prevention Web site. Available at: http://www.cdc.gov/sids. Updated September 4, 2012. Accessed December 31, 2012.

        • Shapiro-Mendoza C.K.
        • Camperlengo L.T.
        • Kim S.Y.
        • Covington T.
        The Sudden Unexpected Infant Death Case Registry: a method to improve surveillance.
        Pediatrics. 2012; 129: e486-e493
        • Shapiro-Mendoza C.K.
        • Kimball M.
        • Tomashek K.M.
        • Anderson R.N.
        • Blanding S.
        US infant mortality trends attributable to accidental suffocation and strangulation in bed from 1984 through 2004: are rates increasing?.
        Pediatrics. 2009; 123: 533-539
        • Willinger M.
        • James L.S.
        • Catz C.
        Defining the sudden infant death syndrome (SIDS): deliberations of an expert panel convened by the National Institute of Child Health and Human Development.
        Fetal Pediatr Pathol. 1991; 11: 677-684
        • Kim S.Y.
        • Shapiro-Mendoza C.K.
        • Chu S.Y.
        • Camperlengo L.T.
        • Anderson R.N.
        Differentiating cause-of-death terminology for deaths coded as sudden infant death syndrome, accidental suffocation, and unknown cause: an investigation using US death certificates, 2003-2004.
        J Forensic Sci. 2012; 57: 364-369
        • Shapiro-Mendoza C.K.
        • Tomashek K.M.
        • Anderson R.N.
        • Wingo J.
        Recent national trends in sudden, unexpected infant deaths: more evidence supporting a change in classification or reporting.
        Am J Epidemiol. 2006; 163: 762-769
        • Task Force on Sudden Infant Death Syndrome
        The changing concept of sudden infant death syndrome: diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk.
        Pediatrics. 2005; 116: 1245-1255
        • Task Force on Sudden Infant Death Syndrome
        SIDS and other sleep-related infant deaths: expansion of recommendations for a safe infant sleeping environment.
        Pediatrics. 2011; 128: e1341-e1367
      2. Kochanek KD, Xu J, Murphy SL, Miniño AM, Hsiang-Ching K. Deaths: final data for 2009. Available at: http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_03.pdf. Published December 29, 2011. Accessed December 31, 2012.

        • Putnam-Hornstein E.
        Report of maltreatment as a risk factor for injury death: a prospective birth cohort study.
        Child Maltreat. 2011; 16: 163-174
        • Jonson-Reid M.
        • Chance T.
        • Drake B.
        Risk of death among children reported for nonfatal maltreatment.
        Child Maltreat. 2007; 12: 86-95
        • O'Halloran R.L.
        • Ferratta F.
        • Harris M.
        • Ilbeigi P.
        • Rom C.D.
        Child abuse reports in families with sudden infant death syndrome.
        Am J Forensic Med Pathol. 1998; 19: 57-62
        • Krous H.F.
        • Haas E.A.
        • Manning J.M.
        • Deeds A.
        • Silva P.D.
        • Chadwick A.E.
        • et al.
        Child protective services referrals in cases of sudden infant death: a 10-year, population-based analysis in San Diego County, California.
        Child Maltreat. 2006; 11: 247-256
        • National Program of Cancer Registries
        Link Plus User's Guide, Version 2.0.
        Centers for Disease Control and Prevention, Atlanta, GA2007
        • Herzog T.N.
        • Scheuren F.J.
        • Winkler W.E.
        Data Quality and Record Linkage Techniques.
        Springer, New York, NY2007
        • World Health Organization
        International statistical classification of diseases and related health problems: Tenth revision.
        World Health Organization, Geneva, Switzerland1992
        • Hussey J.M.
        • Marshall J.M.
        • English D.J.
        • Knight E.D.
        • Lau A.S.
        • Dubowitz H.
        • et al.
        Defining maltreatment according to substantiation: distinction without a difference?.
        Child Abuse Negl. 2005; 29: 479-492
        • Thompson R.
        • Wiley T.R.
        Predictors of re-referral to child protective services: a longitudinal follow-up of an urban cohort maltreated as infants.
        Child Maltreat. 2009; 14: 89-99
        • Malloy M.H.
        • MacDorman M.
        Changes in the classification of sudden unexpected infant deaths: United States, 1992-2001.
        Pediatrics. 2005; 115: 1247-1253
        • Overpeck M.D.
        • Brenner R.A.
        • Cosgrove C.
        • Trumble A.C.
        • Kochanek K.
        • MacDorman M.
        National underascertainment of sudden unexpected infant deaths associated with deaths of unknown cause.
        Pediatrics. 2002; 109: 274-283
        • Haglund B.
        • Cnattingius S.
        Cigarette smoking as a risk factor for sudden infant death syndrome: a population-based study.
        Am J Public Health. 1990; 80: 29-32
        • Trachtenberg F.L.
        • Haas E.A.
        • Kinney H.C.
        • Stanley C.
        • Krous H.F.
        Risk factor changes for sudden infant death syndrome after initiation of Back-to-Sleep campaign.
        Pediatrics. 2012; 129: 630-638
        • Ostfeld B.M.
        • Esposito L.
        • Perl H.
        • Hegyi T.
        Concurrent risks in sudden infant death syndrome.
        Pediatrics. 2010; 125: 447-453
        • Schnitzer P.G.
        • Covington T.M.
        • Dykstra H.K.
        Sudden unexpected infant deaths: sleep environment and circumstances.
        Am J Public Health. 2012; 102: 1204-1212
        • Wisborg K.
        • Kesmodel U.
        • Henriksen T.B.
        • Olsen S.F.
        • Secher N.J.
        A prospective study of smoking during pregnancy and SIDS.
        Arch Dis Child. 2000; 83: 203-206
        • Milerad J.
        • Vege A.
        • Opdal S.H.
        • Rognum T.O.
        Objective measurements of nicotine exposure in victims of sudden infant death syndrome and in other unexpected child deaths.
        J Pediatr. 1998; 133: 232-236
        • Cleves M.
        • Gutierrez R.G.
        • Gould W.
        • Marchenko Y.V.
        An Introduction to Survival Analysis Using Stata.
        3rd ed. Stata Press, College Station, TX2010
        • Fine J.P.
        • Gray R.J.
        A proportional hazards model for the subdistribution of a competing risk.
        J Am Stat Assoc. 1999; 94: 496-509
        • StataCorp
        Stata Statistical Software: Release 12.1.
        Stata Press, College Station, TX2012
        • Young N.K.
        • Boles S.M.
        • Otero C.
        Parental substance use disorders and child maltreatment: overlap, gaps, and opportunities.
        Child Maltreat. 2007; 12: 137-149
        • Phillips D.P.
        • Brewer K.M.
        • Wadensweiler P.
        Alcohol as a risk factor for sudden infant death syndrome (SIDS).
        Addiction. 2011; 106: 516-525
        • Kandall S.R.
        • Gaines J.
        Maternal substance use and subsequent sudden infant death syndrome (SIDS) in offspring.
        Neurotoxicol Teratol. 1991; 13: 235-240
        • Durand D.J.
        • Espinoza A.M.
        • Nickerson B.G.
        Association between prenatal cocaine exposure and sudden infant death syndrome.
        J Pediatr. 1990; 117: 909-911
        • Iyasu S.
        • Randall L.L.
        • Welty T.K.
        • Hsia J.
        • Kinney H.C.
        • Mandell F.
        • et al.
        Risk factors for sudden infant death syndrome among Northern Plains Indians.
        JAMA. 2002; 288: 2717-2723
        • Hymel K.P.
        National Association of Medical Examiners. Distinguishing sudden infant death syndrome from child abuse fatalities.
        Pediatrics. 2006; 118: 421-427
        • Blair P.S.
        • Sidebotham P.
        • Berry P.J.
        • Evans M.
        • Fleming P.J.
        Major epidemiological changes in sudden infant death syndrome: a 20-year population-based study in the UK.
        Lancet. 2006; 367: 314-319
        • Leach C.E.
        • Blair P.S.
        • Fleming P.J.
        • Smith I.J.
        • Platt M.W.
        • Berry P.J.
        • et al.
        Epidemiology of SIDS and explained sudden infant deaths.
        Pediatrics. 1999; 104 (Available at:) (Published October 1, 1999. Accessed December 31, 2012): e43
      3. US Department of Health and Human Services. Child maltreatment: 2011. Administration for Children and Families, Administration on Children, Youth, and Families, Children's Bureau Web site. Available at: http://www.acf.hhs.gov/sites/default/files/cb/cm11.pdf. Published December 12, 2012. Accessed December 31, 2012.