A comparison of cases from 1983-1992 with those from 1993-2006.

* There was no protocol telling employees to check behind them when passing through secured doors. The abductor had followed a maintenance worker onto the unit.

* No one reported the presence of a suspicious person, although at least one person had seen the woman changing clothes.

* Neither the unit alarm nor the door locks worked properly.

* demographic data on 119 NCMEC cases.

* FBI interviews with 14 abductors.

* FBI data on an additional 48 abductors.

* University of Pennsylvania School of Nursing faculty interviews with 38 family members of abducted infants.

* outcomes data on 72 criminal-court cases obtained through telephone interviews with criminal justice system staff.

Sample. This report includes the 247 cases in the NCMEC's database at the time of comparison: 121 cases from 1983 through 1992 and 126 cases from 1993 through 2006. (After the original study was completed, two additional cases were reclassified as nonfamily abductions, bringing the total from 119 to 121 in the earlier period.) We compared unequal time periods in order to take into account the impact of the NCMEC guidelines on preventing infant abductions, which were published in 1993 and are discussed below (see Preventive policy, page 38).

There is no law requiring parents; guardians; or health care, social services, or law enforcement personnel to report missing children to the NCMEC. That said, the NCMEC learns of cases in several ways. Some are reported to the NCMEC from the abduction site by law enforcement personnel, clinical staff, or families. Some are reported by other agencies, including the International Association for Healthcare Security and Safety and the FBI's National Center for the Analysis of Violent Crime. And some come from health care professionals who hear of a case in the news media. The NCMEC also monitors news accounts of abductions.

All nonfamily infant abductions, occurring in the United States and reported to the NCMEC were reviewed and coded according to the following variables:

* infant: age, race, sex, year of abduction, and whether the infant was injured

* parent: whether the infant was taken from its mother, father, or guardian; whether the parent was injured; and whether a weapon was used against the parent

* location of the abduction

* abductor: age, sex, race, employment status, prior arrests, obstetric history, marital and parental status; whether impersonation or disguise was used; whether the abductor's partner was aware of the abduction; whether an accomplice was involved; the abductor's explanation of motive; and whether the abductor was arrested, indicted, and sentenced

* recovery of infant: location, time to recovery, source of tip, and positive identification of the child

Location of abduction. One hundred sixteen infants (47%) were taken from hospitals and other health care settings such as birth centers and prenatal or pediatrics clinics. Another 131 infants (53%) were taken from non-health care locations: 97 (74%) from homes and 34 (26%) from "other" places.

Infants. Roughly equal numbers of infant boys (n = 126) and girls (n = 121) were abducted; 103 infants (42%) were black, 75 (30%) were white, 61 (25%) were Hispanic, and eight (3%) were from other races. The vast majority of abducted infants (95%) were found; 12 (5%) remain missing.

Abductors. The vast majority (237 of the 247 abductors) were women (96%); six were men (2%); the abductor's sex was not recorded in four cases (2%). Ninety-nine (41%) were black, 93 (39%) were white, and 47 (20%) were Hispanic; in eight cases (3%) race wasn't recorded. (We defined race according to the categories that law enforcement agencies used at the time, which were based on definitions from the National Crime Information Center. Many agencies now define Hispanic as an ethnic, rather than racial, category.)

The abductors ranged in age from 14 to 53 years: 90 (36%) were in their 20s, 64 (26%) were in their 30s, 49 (20%) were in their teens, 24 (10%) were in their 40s, and three (1%) were 50 years of age or older. Age was not recorded in 17 (7%) cases. Of the 156 cases with data on the abductor's obstetric history and parental status, 69 (44%) had living biologic children, 27 (17%) reported having had a miscarriage or a stillbirth, and 11 (7%) had both living children and a history of miscarriage. Ninety-one (37%) of the 247 cases were missing these data.

Abductors were known to have made a prior visit to the abduction site in 157 (75%) of the 210 cases reporting this information. One hundred ten abductors (45%) impersonated a health care worker during the crime, and 33 (13%) had made a prior abduction attempt. (We considered impersonation and disguise separately: not all those who impersonated health care or social services personnel wore a disguise such as a uniform, and not all those who wore disguises attempted impersonation.)

Forty-four abductors (18%) used violence, which occurred in nine (8%) of the 116 abductions at health care facilities and in 35 (27%) of the 131 abductions at non-health care locations. (See Abductions Involving Violence, at left.)

SIGNIFICANT FINDINGS: 1983-1992 VS. 1993-2006

Abductor marital status. From the earlier to the later period, the percentage of suspects who were married dropped from 28% to 19%, while the percentage of those who were unmarried increased from 32% to 52%. The percentage of divorced suspects remained at 2%. (In the remaining cases, marital status was unknown.) See Table 2, online.

Time to recovery of the infant decreased significantly from the earlier to the later period. The proportion of cases in which recovery happened within 12 hours of the abduction increased from 20% to 27%; that in which recovery happened within 12 to 24 hours increased from 30% to 36%. Infants were not recovered in 6% of the earlier cases and in 4% of the later cases. See Table 3, online.

When the relationship between time to recovery of the infant and location of the abduction was considered, the data show that, since 1993, infants taken from health care facilities have been recovered faster than those taken from other locations. The percentage of infants taken from health care facilities and recovered within 24 hours increased from 57% in the earlier period to 85% in the later period. Infants abducted from non-health care locations didn't fare as well; in both the earlier and later periods, a smaller percentage of infants was recovered within 24 hours (38% and 52%, respectively). See Table 4, online.

Disguise and impersonation. Disguise, which sometimes involved wearing a uniform, was used significantly more often in health care facility-based abductions during the earlier period, when 41% of abductors did so, compared with 11% in the later period. Similarly, the proportion of abductors who impersonated a health care worker also dropped markedly, from 60% in the earlier period to 29% in the later one. See Tables 5 and 6, online.

Figure 5. Weapon Used: Changes Between Time Periods*

ADDITIONAL FINDINGS: 1983-1992 VS. 1993-2006

Some other findings are worth noting, although these were not statistically significant.

Abductor's race. The proportion of Hispanic abductors nearly doubled, from 14% to 25%; that of black abductors also increased, although less dramatically, from 40% to 43%. The proportion of white abductors decreased from 46% to 32%. See Table 7, online.

Source of tip. The proportion of cases in which family members reported the abductor to law-enforcement authorities increased from 12% to 17%. The proportion in which news reports resulted in tips from the general public to law enforcement authorities increased from 20% to 30%. See Table 8, online.

Arrests and indictments. The proportion of suspects arrested increased from 88% to 94%; the proportion indicted increased from 87% to 94%. See Table 9, online.

Partner awareness. The term partner refers to the abductor's significant other, whether or not the two were married or cohabiting. The data suggest that partners in the later period were less aware of the abductors' actions than those in the earlier period. The percentage of partners who were aware that the abductor had taken an infant declined, from 17% to 10%, while the percentage of those who were unaware increased from 40% to 47%. The percentage of cases in which partner awareness was unclear remained stable at 44%. See Table 10, online.

Prior visit. A greater percentage of abductors made a prior visit to the abduction location in the later period (80%) than in the earlier period (69%). See Table 11, online.

WHAT NURSES NEED TO KNOW

Regarding the site of abduction, a trend was seen away from health care facilities and toward homes, which may be reflected in a rise in parental injury and a greater use of weapons during abduction (perhaps abductors anticipated stronger opposition from parents than from health care facility staff). Abductors' prior visits to abduction sites increased in the later period perhaps because factors such as physical layout and the presence of others are less predictable in non-health care settings, and it may have become more important for abductors to familiarize themselves with locations ahead of time.

The increased numbers of suspects arrested and indicted in the later period may be due to heightened public awareness of nonfamily infant abduction. And the decline in partner awareness may be related to a decline in the number of abductors who were married or cohabiting or both.

Our data suggest some differences in the profiles of nonfamily abductors who take infants from health care settings and those who do so from non-health care settings. See Nonfamily Infant Abductor Profiles, at right.

Parent education should include familiarizing expectant and new parents with their hospital or birthing facility's protocols concerning routine care, visitation, and security. They should be cautioned to be watchful of anyone they don't know well, especially anyone whom they met during the pregnancy, anyone who shows excessive interest, and anyone who arrives at their home unannounced or who can't provide checkable identification. If an abduction occurs, it's essential that parents contact local law enforcement immediately; all such incidents should be reported to the NCMEC as well.

Further research is needed in nonfamily infant abduction, so that more specific preventive guidelines can be developed. Additional study of abductors could provide insight into their characters, thinking patterns, and motivations. We were particularly troubled by the higher incidence of violent abductions; research might help clarify the reasons for this increase. Follow-up on the impact of infant abduction on families is critical.

Abductions Involving Violence

Because there were relatively few data on abductions that involved violence, we considered all cases occurring from 1983 through 2006 together. Violence, defined here as the use of a weapon or force or both against the target infant's parent or guardian, occurred in 44 (18%) of the 247 cases. Forty-two of the violent abductors were women; two were men. Eighteen already had biologic children. Twenty-nine of the violent abductors took an infant from a residence, six did so from a public place, and nine did so from hospitals or other health care settings. Sixteen of the violent cases (36%) involved accomplices: the accomplice was a man in 12 cases and a woman in four. Five of the violent abductors have never been caught and three of the children are still missing.

* is a woman between the ages of 12 and 50 and is often overweight.

* is compulsive and frequently uses manipulation and deceit to gain access.

* often indicates she has either lost a baby or cannot have one.

* lives in or is familiar with the community in which the abduction occurs.

* can provide good care for the baby once the abduction occurs.

* is often married or living with a man.

* visits nursery and maternity units at more than one facility before the abduction and asks detailed questions about procedures and unit layout; the abductor "may also try to abduct from the home setting."

* usually plans the abduction but doesn't necessarily target a specific infant; often seizes any opportunity that presents.

* frequently impersonates a nurse or other health care provider.

* often becomes familiar with facility staff and work routines, as well as with parents of intended victims.

The characteristics of nonfamily infant abductions from non-health care settings differ somewhat. The first five characteristics listed above still apply. Our data suggest four new characteristics, although only the first demonstrates statistical significance. We believe the typical abductor in the non-health care setting also

* is more likely to be single while claiming to have a partner.

* often targets a mother-whom she may find by visiting hospitals-and tries to meet the target family.

* often both plans the abduction and brings a weapon, although the weapon might not be used.

* often impersonates a health care or social services professional when visiting the home.

REFERENCES